McKinsey Global Institute May 2013 Disruptive technologies: Advances that will transform life, business, and the global economy James Manyika Michael Chui Jacques Bughin Richard Dobbs Peter Bisson Alex Marrs
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy Preface Technology is moving so quickly, and in so many directions, that it becomes chal enging to even pay attention—we are victims of “next new thing” fatigue. Yet technology advancement continues to drive economic growth and, in some cases, unleash disruptive change. Economical y disruptive technologies—like the semiconductor microchip, the Internet, or steam power in the Industrial Revolution—transform the way we live and work, enable new business models, and provide an opening for new players to upset the established order. Business leaders and policy makers need to identify potential y disruptive technologies, and careful y consider their potential, before these technologies begin to exert their disruptive powers in the economy and society. In this report, the McKinsey Global Institute (MGI) assesses the potential reach and scope, as wel as the potential economic impact and disruption of major rapidly advancing technology areas. Through extensive research, we sort through the noise to identify 12 technology areas with the potential for massive impact on how people live and work, and on industries and economies. We also attempt to quantify the potential economic impact of each technology across a set of promising applications in 2025. By definition such an exercise is incomplete—technology and innovations always surprise. The potential applications we consider reflect what McKinsey experts and respected leaders in industry and academia who aided our research believe are il ustrative of emerging applications over the next decade or two and provide a good indication of the size and shape of the impact that these applications could have. The combined potential economic impact by 2025 from the applications of the 12 technologies that we have sized may be denominated in the tens of tril ions of dol ars per year. Some of this economic potential wil end up as consumer surplus; a substantial portion of this economic potential wil tranlate into new revenue that companies wil capture and that wil contribute to GDP growth. Other effects could include shifts in profit pools between companies and industries. Our goal in pursuing this research is not to make predictions, either about the specific applications or the specific sizes of impact. Rather we hope this report wil act as a guide for leaders to use as they consider the reach and scope of impact, as wel as the types of impacts that these disruptive technologies could have for the growth and performance of their organizations. We ful y expect and hope others wil build on and enrich this research, as we plan to do. As a companion piece to this research on disruptive technologies, we have updated prior work on business trends enabled by information technologies, which wil be available for download at the MGI website (www.mckinsey.com/mgi). In any case, we believe that these technologies wil have large and disruptive impact. More importantly, the results of our research show that business leaders and policy makers—and society at large—wil confront change on many fronts: in the way businesses organize themselves, how jobs are defined, how we use technology to interact with the world (and with each other), and, in the case of
next-generation genomics, how we understand and manipulate living things. There wil be disruptions to established norms, and there wil be broad societal chal enges. Nevertheless, we see considerable reason for optimism. Many technologies on the horizon offer immense opportunities. We believe that leaders can seize these opportunities, if they start preparing now. This work was led by James Manyika, an MGI director in San Francisco, and Michael Chui, an MGI principal, working closely with McKinsey directors Jacques Bughin and Peter Bisson. We are particularly indebted to our team leaders—Alex Marrs, who managed the project, and Joi Danielson, who co- led a portion of the research. The project team included Hyungpyo Choi, Shalabh Gupta, Tim Wegner, Angela Winkle, and Sabina Wizander. Geoffrey Lewis provided editorial support, Karla Arias assisted with research, and we thank the MGI production and communication staff: Marisa Carder, Julie Philpot, Gabriela Ramirez, and Rebeca Robboy. We thank McKinsey experts whose insight and guidance were critical to our work, in particular directors Matt Rogers on oil and gas exploration and recovery, Stefan Heck on renewable energy, Philip Ma on genomics, and Mona Mourshed on education and training. Roger Roberts, a principal in our Business Technology Office provided multiple insights across various areas of technology. Katy George, a director in the North American Operations Practice, provided expertise on manufacturing, as did Lou Rassey, a principal in the practice. Susan Lund, an MGI principal, provided insight on the changing nature of work and on energy. We were assisted by many experts in our Business Technology Office, including Steve Cheng and Brian Milch, as wel as Bryan Hancock from the Public Sector Practice and Jimmy Sarakatsannis from the Social Sector Practice. We also thank Jonathan Ablett, Sree Ramaswamy, and Vivien Singer for their help on many topics. We are grateful to our external advisers Hal R. Varian, chief economist at Google and emeritus professor in the School of Information, the Haas School of Business and the Department of Economics at the University of California at Berkeley; Erik Brynjolfsson, Schussel Family professor of management at the MIT Sloan School of Management, director of the MIT Center for Digital Business, and research associate at the National Bureau of Economic Research; and Martin Baily, senior fel ow in the Economic Studies Program and Bernard L. Schwartz Chair in Economic Policy Development at the Brookings Institution. This work benefited from the insight of many technology and business thought leaders, including Chamath Palihapitiya, founder and managing partner of The Social+Capital Partnership and a former Facebook executive; Eric Schmidt, executive chairman of Google; and Padmasree Warrior, chief technology and strategy officer of Cisco. We also thank McKinsey alumni Jennifer Buechel, Rob Jenks, and David Mann, as wel as Vivek Wadhwa, vice president of innovation and research at Singularity University; and Ann Winblad, managing director of Hummer Winblad Venture Partners. We thank David Kirkpatrick, CEO of Techonomy; and Paddy Cosgrave, founder of F.ounders. McKinsey col eagues from many practice areas gave generously of their time and expertise to guide our analyses for each technology. Dan Ewing, Ken Kaji , Christian Kraus, Richard Lee, Fredrik Lundberg, Daniel Pacthod, and Remi Said
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy were our experts in mobile Internet technology. For automation of knowledge work, we received input from Rickard Carlsson, Alex Ince-Cushman, Alex Kazaks, Nathan Marston, and Chad Wegner. Our McKinsey experts on the Internet of Things were Mike Greene, Aditi Jain, Nakul Narayan, Jeffrey Thompson, and Peter Weed. Cloud computing insights were provided by Brad Brown, Abhijit Dubey, Loralei Osborn, Naveen Sastry, Kara Sprague, Irina Starikova, and Paul Wil mott. Peter Groves, Craig Melrose, Murali Naidu and Jonathan Til ey provided expertise on advanced robotics. For our research on next-generation genomics, we cal ed on Myoung Cha, Nicolas Denis, Lutz Goedde, Samarth Kulkarni, Derek Neilson, Mark Patel, Roberto Paula, and Pasha Sarraf. In autonomous and near-autonomous vehicles, our experts were Nevin Carr, Matt Jochim, Gustav Lindström, Cody Newman, John Niehaus, and Benno Zerlin. For expertise on energy storage, we cal ed on McKinsey experts Jeremiah Connol y, Mark Faist, Christian Gschwandtner, Jae Jung, Colin Law, Michael Linders, Farah Mandich, Sven Merten, John Newman, Octavian Puzderca, Ricardo Reina, and Kyungyeol Song. In 3D printing, Bartek Blaicke, Tobias Geisbüsch, and Christoph Sohns provided expertise. Helen Chen, Nathan Flesher, and Matthew Veves were our experts on advanced materials. For expert insight on advanced oil and gas exploration and recovery, we relied on Abhijit Akerkar, Drew Erdmann, Bob Frei, Sara Hastings-Simon, Peter Lambert, El en Mo, Scott Nyquist, Dickon Pinner, Joe Quoyeser, Occo Roelofsen, Wombi Rose, Ed Schneider, Maria Fernanda Souto, and Antonio Volpin. In renewable energy, our experts were Ian Banks, Joris de Boer, Sonam Handa, Yunzhi Li, Jurriaan Ruys, Raman Sehgal, and Johnson Yeh. This report is part of our ongoing work about the impact of technology on the economy. Our goal is to provide the fact base and insights about important technological developments that wil help business leaders and policy makers develop appropriate strategies and responses. As with al of MGI’s work, this report has not been sponsored in any way by any business, government, or other institution. Richard Dobbs Director, McKinsey Global Institute Seoul James Manyika Director, McKinsey Global Institute San Francisco May 2013
Connecting rate of improvement and reach today … $5 mil ion vs. $400 Price of the fastest supercomputer in 19751 and an iPhone 4 with equal performance 230+ mil ion Knowledge workers in 2012 $2.7 bil ion, 13 years Cost and duration of the Human Genome Project, completed in 2003 300,000 Miles driven by Google’s autonomous cars with only one accident (human error) 3x Increase in efficiency of North American gas wel s between 2007 and 2011 85% Drop in cost per watt of a solar photovoltaic cel since 2000 1 For CDC-7600, considered the world’s fastest computer from 1969 to 1975; equivalent to $32 mil ion in 2013 at an average inflation rate of 4.3 percent per year since launch in 1969.
… with economic potential in 2025 2–3 bil ion More people with access to the Internet in 2025 $5–7 tril ion Potential economic impact by 2025 of automation of knowledge work $100, 1 hour Cost and time to sequence a human genome in the next decade2 1.5 mil ion Driver-caused deaths from car accidents in 2025, potentially addressable by autonomous vehicles 100–200% Potential increase in North American oil production by 2025, driven by hydraulic fracturing and horizontal dril ing 16% Potential share of solar and wind in global electricity generation by 20253 2 Derek Thompson, “IBM’s kil er idea: The $100 DNA-sequencing machine,” The Atlantic, November 16, 2011. 3 Assuming continued cost declines in solar and wind technology and policy support for meeting the global environmental target of CO concentration lower than 450 ppm by 2050. 2
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 1 Executive summary The parade of new technologies and scientific breakthroughs is relentless and is unfolding on many fronts. Almost any advance is bil ed as a breakthrough, and the list of “next big things” grows ever longer. Yet some technologies do in fact have the potential to disrupt the status quo, alter the way people live and work, rearrange value pools, and lead to entirely new products and services. Business leaders can’t wait until evolving technologies are having these effects to determine which developments are truly big things. They need to understand how the competitive advantages on which they have based strategy might erode or be enhanced a decade from now by emerging technologies—how technologies might bring them new customers or force them to defend their existing bases or inspire them to invent new strategies. Policy makers and societies need to prepare for future technology, too. To do this wel , they wil need a clear understanding of how technology might shape the global economy and society over the coming decade. They wil need to decide how to invest in new forms of education and infrastructure, and figure out how disruptive economic change wil affect comparative advantages. Governments wil need to create an environment in which citizens can continue to prosper, even as emerging technologies disrupt their lives. Lawmakers and regulators wil be chal enged to learn how to manage new biological capabilities and protect the rights and privacy of citizens. Many forces can bring about large-scale changes in economies and societies— demographic shifts, labor force expansion, urbanization, or new patterns in capital formation, for example. But since the Industrial Revolution of the late 18th and early 19th centuries, technology has had a unique role in powering growth and transforming economies. Technology represents new ways of doing things, and, once mastered, creates lasting change, which businesses and cultures do not “unlearn.” Adopted technology becomes embodied in capital, whether physical or human, and it al ows economies to create more value with less input. At the same time, technology often disrupts, supplanting older ways of doing things and rendering old skil s and organizational approaches irrelevant. These economical y disruptive technologies are the focus of our report.1 We view technology both in terms of potential economic impact and capacity to disrupt, because we believe these effects go hand-in-hand and because both are of critical importance to leaders. As the early 20th-century economist Joseph Schumpeter observed, the most significant advances in economies are often accompanied by a process of “creative destruction,” which shifts profit pools, rearranges industry structures, and replaces incumbent businesses. This process is often driven by technological innovation in the hands of entrepreneurs. Schumpeter describes how the Il inois Central railroad’s high-speed freight 1 Recent reports by the McKinsey Global Institute have analyzed how changes in labor forces, global financial markets, and infrastructure investment wil shape economies and influence growth in coming years. See, for example, The world at work: Jobs, pay, and skil s for 3.5 bil ion people, McKinsey Global Institute, June 2012.
2 service enabled the growth of cities yet disrupted established agricultural businesses. In the recent past, chemical-based photography—a technology that dominated for more than a century and continued to evolve—was routed by digital technology in less than 20 years. Today the print media industry is in a life-and-death struggle to remain relevant in a world of instant, online news and entertainment. Some economists question whether technology can stil deliver the kind of wide-ranging, profound impact that the introduction of the automobile or the semiconductor chip had, and point to data showing slowing productivity growth in the United States and the United Kingdom—often early adopters of new technology—as evidence. While we agree that significant chal enges lie ahead, we also see considerable reason for optimism about the potential for new and emerging technologies to raise productivity and provide widespread benefits across economies. Achieving the ful potential of promising technologies while addressing their chal enges and risks wil require effective leadership, but the potential is vast. As technology continues to transform our world, business leaders, policy makers, and citizens must look ahead and plan. Today, we see many rapidly evolving, potential y transformative technologies on the horizon—spanning information technologies, biological sciences, material science, energy, and other fields. The McKinsey Global Institute set out to identify which of these technologies could have massive, economical y disruptive impact between now and 2025. We also sought to understand how these technologies could change our world and how leaders of businesses and other institutions should respond. Our goal is not to predict the future, but rather to use a structured analysis to sort through the technologies with the potential to transform and disrupt in the next decade or two, and to assess potential impact based on what we can know today, and put these promising technologies in a useful perspective. We offer this work as a guide for leaders to anticipate the coming opportunities and changes. IDENTIFYING THE TECHNOLOGIES THAT MATTER The noise about the next big thing can make it difficult to identify which technologies truly matter. Here we attempt to sort through the many claims to identify the technologies that have the greatest potential to drive substantial economic impact and disruption by 2025 and to identify which potential impacts leaders should know about. Important technologies can come in any field or emerge from any scientific discipline, but they share four characteristics: high rate of technology change, broad potential scope of impact, large economic value that could be affected, and substantial potential for disruptive economic impact. Many technologies have the potential to meet these criteria eventual y, but leaders need to focus on technologies with potential impact that is near enough at hand to be meaningful y anticipated and prepared for. Therefore, we focused on technologies that we believe have significant potential to drive economic impact and disruption by 2025. The technology is rapidly advancing or experiencing breakthroughs. Disruptive technologies typical y demonstrate a rapid rate of change in capabilities in terms of price/performance relative to substitutes and alternative approaches, or they experience breakthroughs that drive accelerated rates of change or discontinuous capability improvements. Gene-sequencing technology, for example, is advancing at a rate even faster than computer
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 3 processing power and could soon make possible inexpensive desktop sequencing machines. Advanced materials technology is experiencing significant breakthroughs, from the first artificial production of graphene (a nanomaterial with extraordinary properties including strength and conductivity) in 2004, to IBM’s creation of the first graphene-based integrated circuit in 2011.2 The potential scope of impact is broad. To be economical y disruptive, a technology must have broad reach--touching companies and industries and affecting (or giving rise to) a wide range of machines, products, or services. The mobile Internet, for example, could affect how 5 bil ion people go about their lives, giving them tools to become potential innovators or entrepreneurs— making the mobile Internet one our most impactful technologies. And the Internet of Things technology could connect and embed intel igence in bil ions of objects and devices al around the world, affecting the health, safety, and productivity of bil ions of people. Significant economic value could be affected. An economical y disruptive technology must have the potential to create massive economic impact. The value at stake must be large in terms of profit pools that might be disrupted, additions to GDP that might result, and capital investments that might be rendered obsolete. Advanced robotics, for example, has the potential to affect $6.3 tril ion in labor costs global y. Cloud has the potential to improve productivity across $3 tril ion in global enterprise IT spending, as wel as enabling the creation of new online products and services for bil ions of consumers and mil ions of businesses alike. Economic impact is potentially disruptive. Technologies that matter have the potential to dramatical y change the status quo. They can transform how people live and work, create new opportunities or shift surplus for businesses, and drive growth or change comparative advantage for nations. Next- generation genomics has the potential to transform how doctors diagnose and treat cancer and other diseases, potential y extending lives. Energy storage technology could change how, where, and when we use energy. Advanced oil and gas exploration and recovery could fuel economic growth and shift value across energy markets and regions. To reach our final list of a dozen economical y disruptive technologies we started with more than 100 possible candidates drawn from academic journals, the business and technology press, analysis of published venture capital portfolios, and hundreds of interviews with relevant experts and thought leaders. We assessed each candidate according to our four criteria, eliminating some that were too narrow and others that seem unlikely to start having significant economic impact within our time period. We believe that the technologies we identify have potential to affect bil ions of consumers, hundreds of mil ions of workers, and tril ions of dol ars of economic activity across industries. The 12 potential y economical y disruptive technologies are listed in Exhibit E1. In Exhibit E2, we show representative metrics of how each technology fulfil s our criteria for speed, range of impact, and potential scale of economic value that could be affected. The values in this chart serve to characterize the broad 2 Yu-Ming Lin et al., “Wafer-scale graphene integrated circuit,” Science, volume 332, number 6035, June 2011.
4 potential of these technologies to drive economic impact and disruption and do not represent our estimates of the potential economic impact by 2025, which we describe in Exhibit E3 below. These numbers are not exhaustive; they are indicative and do not represent al possible applications or potential impacts for each technology.
Exhibit E1 Twelve potentially economically disruptive technologies Mobile Internet Increasingly inexpensive and capable mobile computing devices and Internet connectivity Automation of knowledge Intel igent software systems that can work perform knowledge work tasks involving unstructured commands and subtle judgments Internet of Things Networks of low-cost sensors and actuators for data col ection, monitoring, decision making, and process optimization Cloud technology Use of computer hardware and software resources delivered over a network or the Internet, often as a service Advanced robotics Increasingly capable robots with enhanced senses, dexterity, and intel igence used to automate tasks or augment humans Autonomous and Vehicles that can navigate and operate near-autonomous vehicles with reduced or no human intervention Next-generation genomics Fast, low-cost gene sequencing, advanced big data analytics, and synthetic biology (“writing” DNA) Energy storage Devices or systems that store energy for later use, including batteries 3D printing Additive manufacturing techniques to create objects by printing layers of material based on digital models Advanced materials Materials designed to have superior characteristics (e.g., strength, weight, conductivity) or functionality Advanced oil and gas Exploration and recovery techniques exploration and recovery that make extraction of unconventional oil and gas economical Renewable energy Generation of electricity from renewable sources with reduced harmful climate impact SOURCE: McKinsey Global Institute analysis E1
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 5
Exhibit E2 Speed, scope, and economic value at stake of 12 potentially economically disruptive technologies Illustrative rates of technology improvement Illustrative groups, products, and Illustrative pools of economic value and diffusion resources that could be impacted1 that could be impacted1 Mobile $5 million vs. $4002 4.3 billion $1.7 trillion Internet Price of the fastest supercomputer in 1975 vs. that of People remaining to be connected to the GDP related to the Internet an iPhone 4 today, equal in performance (MFLOPS) Internet, potential y through mobile $25 trillion Internet 6x Interaction and transaction worker Growth in sales of smartphones and tablets since 1 billion employment costs, 70% of global launch of iPhone in 2007 Transaction and interaction workers, employment costs nearly 40% of global workforce Automation 100x 230+ million $9+ trillion of knowledge Increase in computing power from IBM’s Deep Blue Knowledge workers, 9% of global Knowledge worker employment costs, work (chess champion in 1997) to Watson (Jeopardy workforce 27% of global employment costs winner in 2011) 1.1 billion 400+ million Smartphone users, with potential to use Increase in number of users of intel igent digital automated digital assistance apps assistants like Siri and Google Now in last 5 years Internet of 300% 1 trillion $36 trillion Things Increase in connected machine-to-machine devices Things that could be connected to the Operating costs of key affected industries over past 5 years Internet across industries such as (manufacturing, health care, and mining) manufacturing, health care, and mining 80–90% $4 trillion Price decline in MEMS (microelectromechanical 40 million Global health care spend on chronic systems) sensors in last 5 years Annual deaths from chronic diseases diseases like Type 2 diabetes and cardiovascular disease Cloud 18 months 2.7 billion $1.7 trillion technology Time to double server performance per dollar Internet users GDP related to the Internet 3x 50 million $3 trillion Monthly cost of owning a server vs. renting in Servers in the world Enterprise IT spend the cloud Advanced 75–85% 320 million $6 trillion robotics Lower price for Baxter3 than a typical industrial robot Manufacturing workers, 12% of global Manufacturing worker employment costs, workforce 19% of global employment costs 170% Growth in sales of industrial robots, 2009–11 250 million $2–3 trillion Annual major surgeries Cost of major surgeries Autonomous 7 1 billion $4 trillion and near- Miles driven by top-performing driverless car in 2004 Cars and trucks global y Automobile industry revenues autonomous DARPA Grand Challenge along a 150-mile route 450,000 $155 billion vehicles 1,540 Civilian, military, and general aviation Revenue from sales of civilian, military, and Miles cumulatively driven by cars competing in 2005 aircraft in the world general aviation aircraft Grand Chal enge 300,000+ Miles driven by Google’s autonomous cars with only 1 accident (which was human-caused) Next- 10 months 26 million $6.5 trillion generation Time to double sequencing speed per dollar Annual deaths from cancer, cardio- Global health-care costs genomics vascular disease, or Type 2 diabetes 100x $1.1 trillion Increase in acreage of genetical y modified crops, 2.5 billion Global value of wheat, rice, maize, soy, 1996–2012 People employed in agriculture and barley Energy 40% 1 billion $2.5 trillion storage Price decline for a lithium-ion battery pack in an Cars and trucks global y Revenue from global consumption of electric vehicle since 2009 gasoline and diesel 1.2 billion People without access to electricity $100 billion Estimated value of electricity for households currently without access 3D printing 90% 320 million $11 trillion Lower price for a home 3D printer vs. 4 years ago Manufacturing workers, 12% of global Global manufacturing GDP workforce 4x $85 billion Increase in additive manufacturing revenues in past 8 billion Revenue from global toy sales 10 years Annual number of toys manufactured globally Advanced $1,000 vs. $50 7.6 million tons $1.2 trillion materials Difference in price of 1 gram of nanotubes over Annual global silicon consumption Revenue from global semiconductor sales 10 years 45,000 metric tons $4 billion 115x Annual global carbon fiber consumption Revenue from global carbon fiber sales Strength-to-weight ratio of carbon nanotubes vs. steel Advanced 3x 22 billion $800 billion oil and gas Increase in efficiency of US gas wel s between Barrels of oil equivalent in natural gas Revenue from global sales of natural gas exploration 2007 and 2011 produced global y $3.4 trillion and recovery 2x 30 billion Revenue from global sales of crude oil Increase in efficiency of US oil wel s between Barrels of crude oil produced global y 2007 and 2011 Renewable 85% 21,000 TWh $3.5 trillion energy Lower price for a solar photovoltaic cel per watt since Annual global electricity consumption Value of global electricity consumption 2000 13 billion tons $80 billion 19x Annual CO Value of global carbon market transactions 2 emissions from electricity Growth in solar photovoltaic and wind generation generation, more than from al cars, capacity since 2000 trucks, and planes 1 Not comprehensive; indicative groups, products, and resources only. 2 For CDC-7600, considered the world’s fastest computer from 1969 to 1975; equivalent to $32 mil ion in 2013 at an average inflation rate of 4.3% per year since launch in 1969. 3 Baxter is a general-purpose basic manufacturing robot developed by startup Rethink Robotics. SOURCE: McKinsey Global Institute analysis
6 Mobile Internet In just a few years, Internet-enabled portable devices have gone from a luxury for a few to a way of life for more than 1 bil ion people who own smartphones and tablets. In the United States, an estimated 30 percent of Web browsing and 40 percent of social media use is done on mobile devices; by 2015, wireless Web use is expected to exceed wired use. Ubiquitous connectivity and an explosive proliferation of apps are enabling users to go about their daily routines with new ways of knowing, perceiving, and even interacting with the physical world. The technology of the mobile Internet is evolving rapidly, with intuitive interfaces and new formats, including wearable devices. The mobile Internet also has applications across businesses and the public sector, enabling more efficient delivery of many services and creating opportunities to increase workforce productivity. In developing economies, the mobile Internet could bring bil ions of people into the connected world. Automation of knowledge work Advances in artificial intel igence, machine learning, and natural user interfaces (e.g., voice recognition) are making it possible to automate many knowledge worker tasks that have long been regarded as impossible or impractical for machines to perform. For instance, some computers can answer “unstructured” questions (i.e., those posed in ordinary language, rather than precisely written as software queries), so employees or customers without specialized training can get information on their own. This opens up possibilities for sweeping change in how knowledge work is organized and performed. Sophisticated analytics tools can be used to augment the talents of highly skil ed employees, and as more knowledge worker tasks can be done by machine, it is also possible that some types of jobs could become ful y automated. Internet of Things The Internet of Things—embedding sensors and actuators in machines and other physical objects to bring them into the connected world—is spreading rapidly. From monitoring the flow of products through a factory to measuring the moisture in a field of crops to tracking the flow of water through utility pipes, the Internet of Things al ows businesses and public-sector organizations to manage assets, optimize performance, and create new business models. With remote monitoring, the Internet of Things also has great potential to improve the health of patients with chronic il nesses and attack a major cause of rising health-care costs. Cloud With cloud technology, any computer application or service can be delivered over a network or the Internet, with minimal or no local software or processing power required. In order to do this, IT resources (such as computation and storage) are made available on an as-needed basis—when extra capacity is needed it is seamlessly added, without requiring up-front investment in new hardware or programming. The cloud is enabling the explosive growth of Internet-based services, from search to streaming media to offline storage of personal data (photos, books, music), as wel as the background processing capabilities that enable mobile Internet devices to do things like respond to spoken commands to ask for directions. The cloud can also improve the economics of IT for companies and governments, as wel as provide greater flexibility and responsiveness. Final y, the cloud can enable entirely new business models, including al kinds of pay-as- you-go service models.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 7 Advanced robotics For the past several decades, industrial robots have taken on physical y difficult, dangerous, or dirty jobs, such as welding and spray painting. These robots have been expensive, bulky, and inflexible—bolted to the floor and fenced off to protect workers. Now, more advanced robots are gaining enhanced senses, dexterity, and intel igence, thanks to accelerating advancements in machine vision, artificial intel igence, machine-to-machine communication, sensors, and actuators. These robots can be easier for workers to program and interact with. They can be more compact and adaptable, making it possible to deploy them safely alongside workers. These advances could make it practical to substitute robots for human labor in more manufacturing tasks, as wel as in a growing number of service jobs, such as cleaning and maintenance. This technology could also enable new types of surgical robots, robotic prosthetics, and “exoskeleton” braces that can help people with limited mobility to function more normal y, helping to improve and extend lives. Next-generation genomics Next-generation genomics marries advances in the science of sequencing and modifying genetic material with the latest big data analytics capabilities. Today, a human genome can be sequenced in a few hours and for a few thousand dol ars, a task that took 13 years and $2.7 bil ion to accomplish during the Human Genome Project. With rapid sequencing and advanced computing power, scientists can systematical y test how genetic variations can bring about specific traits and diseases, rather than using trial and error. Relatively low-cost desktop sequencing machines could be used in routine diagnostics, potential y significantly improving treatments by matching treatments to patients. The next step is synthetic biology—the ability to precisely customize organisms by “writing” DNA. These advances in the power and availability of genetic science could have profound impact on medicine, agriculture, and even the production of high-value substances such as biofuels—as wel as speed up the process of drug discovery. Autonomous and near-autonomous vehicles It is now possible to create cars, trucks, aircraft, and boats that are completely or partly autonomous. From drone aircraft on the battlefield to Google’s self- driving car, the technologies of machine vision, artificial intel igence, sensors, and actuators that make these machines possible is rapidly improving. Over the coming decade, low-cost, commercial y available drones and submersibles could be used for a range of applications. Autonomous cars and trucks could enable a revolution in ground transportation—regulations and public acceptance permitting. Short of that, there is also substantial value in systems that assist drivers in steering, braking, and col ision avoidance. The potential benefits of autonomous cars and trucks include increased safety, reduced CO emissions, 2 more leisure or work time for motorists (with hands-off driving), and increased productivity in the trucking industry.
8 Energy storage Energy storage technology includes batteries and other systems that store energy for later use. Lithium-ion batteries and fuel cel s are already powering electric and hybrid vehicles, along with bil ions of portable consumer electronics devices. Li-ion batteries in particular have seen consistent increases in performance and reductions in price, with cost per unit of storage capacity declining dramatical y over the past decade. Over the next decade, advances in energy storage technology could make electric vehicles (hybrids, plug-in hybrids, and al -electrics) cost competitive with vehicles based on internal-combustion engines. On the power grid, advanced battery storage systems can help with the integration of solar and wind power, improve quality by control ing frequency variations, handle peak loads, and reduce costs by enabling utilities to postpone infrastructure expansion. In developing economies, battery/solar systems have the potential to bring reliable power to places it has never reached. 3D printing Until now, 3D printing has largely been used by product designers and hobbyists and for a few select manufacturing applications. However, the performance of additive manufacturing machinery is improving, the range of materials is expanding, and prices (for both printers and materials) are declining rapidly— bringing 3D printing to a point where it could see rapid adoption by consumers and even for more manufacturing uses. With 3D printing, an idea can go directly from a 3D design file to a finished part or product, potential y skipping many traditional manufacturing steps. Importantly, 3D printing enables on-demand production, which has interesting implications for supply chains and for stocking spare parts—a major cost for manufacturers. 3D printing can also reduce the amount of material wasted in manufacturing and create objects that are difficult or impossible to produce with traditional techniques. Scientists have even “bioprinted” organs, using an inkjet printing technique to layer human stem cel s along with supporting scaffolding. Advanced materials Over the past few decades, scientists have discovered ways to produce materials with incredible attributes—smart materials that are self-healing or self-cleaning; memory metals that can revert to their original shapes; piezoelectric ceramics and crystals that turn pressure into energy; and nanomaterials. Nanomaterials in particular stand out in terms of their high rate of improvement, broad potential applicability, and long-term potential to drive massive economic impact. At nanoscale (less than 100 nanometers), ordinary substances take on new properties—greater reactivity, unusual electrical properties, enormous strength per unit of weight—that can enable new types of medicine, super-slick coatings, stronger composites, and other improvements. Advanced nanomaterials such as graphene and carbon nanotubes could drive particularly significant impact. For example, graphene and carbon nanotubes could help create new types of displays and super-efficient batteries and solar cel s. Final y, pharmaceutical companies are already progressing in research to use nanoparticles for targeted drug treatments for diseases such as cancer.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 9 Advanced oil and gas exploration and recovery The ability to extract so-cal ed unconventional oil and gas reserves from shale rock formations is a technology revolution that has been gathering force for nearly four decades. The combination of horizontal dril ing and hydraulic fracturing makes it possible to reach oil and gas deposits that were known to exist in the United States and other places but that were not economical y accessible by conventional dril ing methods. The potential impact of this technology has received enormous attention. With continued improvements, this technology could significantly increase the availability of fossil fuels for decades and produce an immediate boon for energy-intensive industries such as petrochemicals manufacturing. Eventual y, improving technology for oil and gas exploration and recovery could even unlock new types of reserves, including coalbed methane, tight sandstones, and methane clathrates (also known as methane hydrates), potential y ushering in another energy “revolution.” Renewable energy Renewable energy sources such as solar, wind, hydro-electric, and ocean wave hold the promise of an endless source of power without stripping resources, contributing to climate change, or worrying about competition for fossil fuels. Solar cel technology is progressing particularly rapidly. In the past two decades, the cost of power produced by solar cel s has dropped from nearly $8 per watt of capacity to one tenth of that amount. Meanwhile, wind power constitutes a rapidly growing proportion of renewable electricity generation. Renewable energy sources such as solar and wind are increasingly being adopted at scale in advanced economies like the United States and the European Union. Even more importantly, China, India, and other emerging economies have aggressive plans for solar and wind adoption that could enable further rapid economic growth while mitigating growing concerns about pol ution. The 12 technologies in our final list do not represent al potential y economical y disruptive technologies in 2025. Many of the other advancing technologies that we reviewed are also worth fol owing and thinking about. In our view they do not have the same potential for economic impact and disruption by 2025, but we cannot rule out sudden breakthroughs or other factors, such as new public policies, that might change that (see Box 1, “Other technologies on the radar”).
10 Box 1. Other technologies on the radar Some of the technologies that we reviewed, but which did not make our final list, are nonetheless interesting and worthy of consideration. Here we note two groups of these technologies. Five technologies that nearly made our final list: Next-generation nuclear (fission) has potential to disrupt the global energy mix but seems unlikely to create significant impact by 2025 given the time frames of current experiments and pilots. Fusion power also has massive potential, but it is even more speculative than next-generation nuclear fission in terms of both technological maturity and time frame. Carbon sequestration could have great impact on reducing CO 2 concentration in the atmosphere, but despite sustained R&D investment it may not become cost-effective and deployed at scale by 2025. Advanced water purification could benefit mil ions of people facing water shortages, but approaches with substantial y better economics than currently known approaches may not be operating at scale by 2025. Quantum computing represents a potential y transformative alternative to digital computers, but the breadth of its applicability and impact remain unclear and the time frame for commercialization is uncertain. A sampling of other interesting and often hyped candidates that were not close in the final running: Private space flight is likely to be limited to space tourism and private satellite launches through 2025, though after that, applications such as asteroid mining could drive greater economic impact. OLED / LED lighting has potential for extensive reach in terms of people affected but seems unlikely to disrupt pools of economic value beyond narrow industries by 2025. Wireless charging is promising for some applications but overal offers limited impact at high cost. Simple versions exist, but it is not clear that the technology serves an important need versus substitutes such as improved energy storage technology. Flexible displays have long been in development and could offer exciting new possibilities for for mobile the designs of mobile devices and TVs, but on their own seem unlikely to have broad-based disruptive impact by 2025. 3D and volumetric displays have received a lot of attention, but it is not clear that these technologies wil drive broad economic impact by 2025.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 11 ESTIMATED POTENTIAL ECONOMIC IMPACT IN 2025 ACROSS SIZED APPLICATIONS Exhibit E3 shows our estimates of the potential economic impact that select applications of each technology could create in 2025 (see Box 2, “Approach to estimating potential economic impact in 2025”). While these estimates are incomplete by definition, the analysis suggests significant potential impact from even a few possible applications. It is important to note, however, that this economic potential should not be equated with market sizes for these technologies. The economic potential wil be captured as consumer surplus as wel as in new revenue and GDP growth as companies commercialize these technologies. For company leaders, it is worth noting the great extent to which Internet-based technologies have tended to shift value to consumers; in our work we see that as much as two-thirds of the value created by new Internet offerings has been captured as consumer surplus. 3 Moreover, our sizing is not comprehensive: we have estimated the potential economic impact in 2025 of applications that we can anticipate today and which appear capable of affecting large amounts of value. But it is impossible to predict al the ways in which technologies wil be applied; the value created in 2025 could be far larger than what we estimate here. Based on our analysis, however, we are convinced that col ectively the potential for our sized technologies and applications is huge: taken together and netting out potential overlaps, we find that they have the potential to drive direct economic impact on the order of $14 tril ion to $33 tril ion per year in 2025. 3 Internet matters: The Net’s sweeping impact on growth, jobs, and prosperity, McKinsey Global Institute, May 2011.
Estimated potential economic impact Range of sized potential Impact from other
of technologies from sized applications economic impacts potential applications (not sized) Low High in 2025, including consumer surplus X–Y $ tril ion, annual Mobile Internet 3.7–10.8 Automation of 5.2–6.7 knowledge work Internet of Things 2.7–6.2 Cloud technology 1.7–6.2 Advanced robotics 1.7–4.5 Autonomous and near- Notes on sizing 0.2–1.9 autonomous vehicles These estimates of economic impact are not comprehensive and include potential direct Next-generation impact of sized applications only. 0.7–1.6 genomics These estimates do not represent GDP or market size (revenue), but rather economic potential, including consumer Energy storage 0.1–0.6 surplus. Relative sizes of technology categories shown here cannot be considered a “ranking” because 3D printing 0.2–0.6 our sizing is not comprehensive. We do not quantify the split or transfer of surplus among or across companies or consumers. Such transfers would depend on Advanced materials 0.2–0.5 future competitive dynamics and business models. These estimates are not directly Advanced oil and gas additive due to partial y 0.1–0.5 overlapping applications and/or exploration and recovery value drivers across technologies. These estimates are not ful y Renewable energy 0.2–0.3 risk- or probability-adjusted. SOURCE: McKinsey Global Institute analysis E3, also BN1
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 13 Box 2. Approach to estimating potential economic impact in 2025 We focus on estimating the potential economic impact of 12 technologies across a set of promising applications, based on feasible scenarios for technology advancement, reach, and resulting productivity or value gains that could be achieved by 2025. We focus on estimating the potential (rather than realized) value in 2025 by assuming that addressable barriers to technology adoption and value creation (such as the need for supporting regulations) can be overcome and that reasonable, necessary investments can be made. Our estimates represent annual value, including consumer surplus, that could be realized in 2025 across sized applications. These estimates are not potential revenue, market size, or GDP impact. We do not attempt to size al of the many possible indirect and fol ow-on effects. We also do not size possible surplus shifts among companies and industries, or between companies and consumers. Final y, our estimates are not adjusted for risk or probability. To estimate the potential direct economic impact of technologies by 2025, we first identify applications and drivers of value for each technology. We then define a scope of potential impact for each application (for example, the operating cost base of an industry where the introduction of a technology might alter costs) which we project forward to 2025 to create a hypothetical base case in which the technology under examination is effectively “frozen” or held constant with no technology progress, diffusion, or additional use. We next consider potential rates of technology diffusion and adoption across the estimated scope of impact for the application, taking into account price/performance improvement. Final y, we estimate potential productivity or value gains from each application that could be achieved across our defined scope of impact by 2025 to determine the potential direct economic impact of the use of the technology for this application. In some cases, we use prior McKinsey research to estimate a portion of the additional surplus that could be created by use of technologies such as the Internet. In the case of advanced oil and gas exploration and recovery and renewable energy, we focus on estimating the value of additional output that could be cost-effectively produced using improved technology. In many cases there could be a lag between the introduction of new technology and its economic impact, owing in part to the need to reconfigure processes to ful y capture benefits. We account for this lag by factoring in structural constraints such as the need for supporting infrastructure, up-front investments (for example, the cost of advanced robots), and prevailing industry investment cycles. We do not take into account less tangible barriers such as cultural resistance or political opposition, as these barriers could potential y be overcome by 2025. We have focused on quantifying the total value from use of each technology because we believe this is a better measure than GDP or other growth accounting metrics for evaluating the potential of a technology to drive transformative impact on people and the economy. GDP, for example, does not include consumer surplus, which is an important portion of the value created from new technology.
14 SOME OBSERVATIONS While we evaluated each technology separately and sized their potential economic impacts independently, we did observe some interesting patterns in the results. These observations reflect common traits among economical y disruptive technologies. Here we examine a set of overarching implications for stakeholders to consider as they plan for the coming decade of economical y disruptive technology. Information technology is pervasive. Most of the technologies on our list are directly enabled, or enhanced, by information technology. Continuing progress in artificial intel igence and machine learning are essential to the development of advanced robots, autonomous vehicles, and in knowledge work automation tools. The next generation of gene sequencing depends highly on improvements in computational power and big data analytics, as does the process of exploring and tapping new sources of oil and natural gas. 3D printing uses computer generated models and benefits from an online design sharing ecosystem. The mobile Internet, Internet of Things, and cloud are themselves information and communications technologies. Information technologies tend to advance very rapidly, often fol owing exponential trajectories of improvement in cost/ performance. Also, information technologies are often characterized by strong network effects, meaning that the value to any user increases as the number of users multiplies. Just as IT creates network effects for users of social media and the mobile Internet, IT-enabled platforms and ecosystems could bring additional value to users of 3D printing or to researchers experimenting with next-generation genomics technology. In a separate report, also released in May 2013, we take a look at how advances in IT are shaping important business trends in the next few years (see Box 3: Ten IT-enabled business trends for the decade ahead). Box 3. IT-enabled business trends We have revisited and updated previous perspectives on IT-enabled business trends that appeared in the McKinsey Quarterly in 2007 and 2010. These trends are powerful ways in which businesses, organizations, and governments can use information technologies to implement strategy, manage people and assets, alter organizational structures, and build new business models. These IT-enabled business trends are already driving pervasive impact across thousands of companies and across sectors. These trends include some of the technologies in this report, such as automation of knowledge work. Some technologies in this report, such as cloud computing, underpin IT-enabled business trends. The report can be downloaded at www.mckinsey.com/mgi.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 15 Combinations of technologies could multiply impact. We see that certain emerging technologies could be used in combination, reinforcing each other and potential y driving far greater impact. For example, the combination of next-generation genomics with advances in nanotechnology has the potential to bring about new forms of targeted cancer drugs. It is possible that the first commercial y available nano-electromechanical machines (NEMS), molecule-sized machines, could be used to create very advanced sensors for wearable mobile Internet devices or Internet of Things applications. And automated knowledge work capabilities could help drive dramatic advances across many areas, including next-generation genomics. Another example of symbiotic development exists between advances in energy storage and renewable energy sources; the ability to store electricity created by solar or wind helps to integrate renewables into the power grid. The advances in energy storage that make this possible could benefit, in turn, from advances in nanomaterials for batteries. Similarly, the mobile Internet might never live up to its enormous potential without important advances in cloud computing to enable applications—including tools for automating knowledge work—on mobile devices. Consumers could win big, particularly in the long run. Many of the technologies on our list have the potential to deliver the lion’s share of their value to consumers, even while providing producers with sufficient profits to encourage technology adoption and production. Technologies like next- generation genomics and advanced robotics could deliver major health benefits, not al of which may be usable by health-care payers and providers, many of whom face growing pressure to help improve patient outcomes while also reducing health-care costs. Many technologies wil also play out in fiercely competitive consumer markets—particularly on the Internet, where earlier McKinsey research has shown consumers capture the majority of the economic surplus created.4 Mobile Internet, cloud, and the Internet of Things are prime examples. Also, as technologies are commercialized and come into widespread use, competition tends to shift value to consumers. The nature of work will change, and millions of people will require new skills. It is not surprising that new technologies make certain forms of human labor unnecessary or economical y uncompetitive and create demand for new skil s. This has been a repeated phenomenon since the Industrial Revolution: the mechanical loom marginalized home weaving while creating jobs for mill workers. However, the extent to which today’s emerging technologies could affect the nature of work is striking. Automated knowledge work tools will almost certainly extend the powers of many types of workers and help drive top-line improvements with innovations and better decision making, but they could also automate some jobs entirely. Advanced robotics could make more manual tasks subject to automation, including in services where automation has had less impact until now. Business leaders and policy makers wil need to find ways to realize the benefits of these technologies while creating new, innovative ways of working and providing new skil s to the workforce. 4 Internet matters: The Net’s sweeping impact on growth, jobs, and prosperity, McKinsey Global Institute, May 2011.
16 The future for innovators and entrepreneurs looks bright. A new wave of unprecedented innovation and entrepreneurship could be in the offing as a result of fal ing costs and rapid dissemination of technologies. Many of the technologies discussed in this report wil be readily available and may require little or no capital investment. 3D printing, for example, could help “democratize” the design, production, and distribution of products and services. Cloud-based services and mobile Internet devices could help level the playing field, putting IT capabilities and other resources within reach of smal enterprises, including in developing nations. Final y, the opportunities and innovation unleashed by a new wave of entrepreneurship could provide new sources of employment. Technology impact differs between advanced and developing economies. There are many examples: in advanced economies and in the fastest-growing developing economies, the chief value of energy storage could be to make electric vehicles competitive with cars that rely solely on internal- combustion engines. But in the poorest developing economies, advanced batteries can provide millions of people with access to electricity, enabling them to connect to the digital world and join the global economy (Exhibit E4). Advanced robots could be a boon to manufacturing, but could reduce global demand for the low-cost labor that developing economies offer the world and which drives their development. Mobile Internet devices could deliver remarkable new capabilities to many people in advanced economies, but could connect two bil ion to three bil ion more people to the digital economy in the developing world. Also, with less legacy infrastructure and fewer investments in old technology, developing economies could leapfrog to more efficient and capable technologies (e.g., adopting the mobile Internet before telephone or cable-TV wiring has been instal ed, or possibly even adopting solar power plus energy storage solutions before being connected to the power grid). Benefits of technologies may not be evenly distributed. While each of the technologies on our list has potential to create significant value, in some cases this value may not be evenly distributed, and could even contribute to widening income inequality. As MIT economist Erik Brynjolfsson has observed, it is possible that advancing technology, such as automation of knowledge work or advanced robotics, could create disproportionate opportunities for some highly skil ed workers and owners of capital while replacing the labor of some less skil ed workers with machines. This places an even greater importance on training and education to refresh and upgrade worker skil s and could increase the urgency of addressing questions on how best to deal with rising income inequality. The link between hype and potential is not clear. Emerging technologies often receive a great deal of notice. News media know that the public is fascinated with gadgets and eager for information about how the future might unfold. The history of technology is littered with breathless stories of breakthroughs that never quite materialized. The hype machine can be equal y misleading in what it chooses to ignore. As Exhibit E5 shows, with the exception of the mobile Internet, there is no clear relationship between the amount of talk a technology generates and its potential to create value. The lesson for leaders is to make sure that they and their advisers have the knowledge to make their own assessments based on a structured analysis involving multiple scenarios of technology advancement and potential impact.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 17
Exhibit E4 Estimated distribution of potential economic impact between developed and Impact on developing economies for sized applications Developed economies % of potential economic impact for sized applications Developing economies Developed Developing High-value applications, e.g., Bulk of new mobile users Mobile Internet 50 50 increasing worker productivity Automation of Higher impact of increasing labor Large number of knowledge 80 20 knowledge work productivity workers Major applications enabled by Large applicable spend base, advanced technology lower initial adoption Internet of Things 70 30 infrastructure, e.g., advanced supply chain systems Cloud technology 30 70 Higher surplus per user Majority of new adoption Greater ability to pay for surgical Many manufacturing workers but Advanced robotics 80 20 robots and prosthetics; high lower savings from automation savings from automation Autonomous and Early adoption in high-end Many vehicles but smal er near-autonomous 80 20 vehicles percentage of high-end vehicles vehicles and low cost of hiring drivers Next-generation 80 20 Greater early adoption of genomic Lower initial adoption, particularly genomics technologies and treatments for new treatments Many new vehicles with potentially Many vehicles but potential y Energy storage 60 40 higher adoption of electric and smal er percentage of new electric hybrid models and hybrid models Potential for earlier adoption in Large manufacturing base and 3D printing 60 40 manufacturing and by consumers many consumers, but lower initial adoption Greater early adoption of new Lower initial adoption for new nano-based treatments due to nano-based treatments and Advanced materials 90 10 more advanced healthcare substances systems Advanced oil and North America leads in shale gas Significant investments being gas exploration 80 20 and light tight oil production made but could require years to and recovery catch up Larger existing renewables base Large renewables capacity Renewable energy 20 80 (especial y wind) with moderate development, e.g., in China growth Notes on sizing
These economic impact estimates are not comprehensive and include We do not quantify the split or transfer of surplus among or across potential direct impact of sized applications only. companies or consumers, as this would depend on emerging competitive These estimates do not represent GDP or market size (revenue), but dynamics and business models. rather economic potential, including consumer surplus. These estimates are not directly additive due to partial y overlapping Relative sizes of technology categories shown here cannot be applications and/or value drivers across technologies. considered a “ranking” because our sizing is not comprehensive. These estimates are not ful y risk- or probability-adjusted. SOURCE: McKinsey Global Institute analysis E4
Exhibit E5 The relationship between hype about a technology and its
potential economic impact is not clear Media attention Number of relevant articles in major general interest and business publications over 1 year (log scale) 100,000 Mobile Internet Renewable 10,000 energy Advanced oil and Autonomous and gas exploration near-autonomous and recovery vehicles Cloud technology Next- Energy generation storage genomics 1,000 Internet of Things Advanced 3D printing robotics Advanced Automation of materials knowledge work 100 100 1,000 10,000 Potential economic impact across sized applications $ bil ion (log scale) NOTE: Estimates of potential economic impact are for only some applications and is not a comprehensive estimate of total potential impact. Estimates include consumer surplus and cannot be related to potential company revenue, market size, or GDP impact. We do not size possible surplus shifts among companies and industries, or between companies and consumers. These estimates are not risk- or probability-adjusted. SOURCE: Factiva; McKinsey Global Institute analysis Scientific discovery and innovation will surprise us. We examined many technologies to evaluate their potential, but in doing so we were impressed by the reality that it is impossible to predict how new technologies wil emerge and play out. Many of the technologies on our list likely wil , at some point, be revolutionized by advancements in science. The technologies that define the 20th and 21st centuries, including modern medicine and electronics, were enabled by scientific breakthroughs like germ theory and Maxwel ’s laws of electromagnetism. Emerging technologies like genomics and nanotechnology are likewise being driven by unpredictable scientific breakthroughs, from the completion of the Human Genome Project in 2003 to the first artificial production of graphene in 2004. Harnessing the ful potential of advanced nanomaterials like graphene wil require major improvements or breakthroughs in cost-effective production techniques. Moreover, when breakthroughs in technologies like advanced materials or energy storage occur, they could drive impact across a host of applications and sectors, likely including some major direct impacts, but potential y also including a wide array of indirect and fol ow-on impacts. There are some troubling challenges. The technologies on our list have great potential to improve the lives of bil ions of people. Cloud computing and the mobile Internet, for example, could raise productivity and quality in education, health care, and p E5 ublic services. At the same time, some of these technologies could bring unwanted side effects. The benefits of the mobile Internet and cloud computing are accompanied by rising risks of security and privacy breaches. Objects and machines under the control of computers across the Web (the Internet of Things) can also be hacked, exposing
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 19 factories, refineries, supply chains, power plants, and transportation networks to new risks. Next-generation genomics has the potential to grant new powers over biology, but these powers could be abused to disastrous effect. Low-cost desktop gene-sequencing machines wil not only put the power of genomics in doctor offices, but also potential y in the hands of terrorists. Even wel - intentioned experiments in garages using inexpensive sequencing and DNA synthesis equipment could result in the production and release of dangerous organisms. And nanomaterials offer great promise, but more research will be required to ful y ascertain their potential impact on health. It wil be up to business leaders, policy makers, and societies to weigh these risks and navigate a path that maximizes the value of these technologies while avoiding their dangers. IMPLICATIONS As we conducted our research and created estimates of the potential economic impact of disruptive technologies, we focused on identifying how each of these technologies could affect individuals, societies, organizations, economies, and governments in transformative and disruptive ways. Exhibit E6 lays out some major ways in which each technology on our list could drive transformative and disruptive impact by 2025. In considering the disruptive potential of these technologies, we see that each could drive profound changes across many dimensions—in the lives of citizens, in business, and across the global economy. As noted, the future seems bright for entrepreneurs and innovators. 3D printing, the mobile Internet, cloud technology, and even next-generation genomics could provide the opportunities and the tools to al ow smal enterprises to compete on a meaningful scale and advance into new markets rapidly. Many technologies, including advanced robotics, next-generation genomics, and renewable energy, have real potential to drive tangible improvements in quality of life, health, and the environment. For example, advanced robotic surgical systems and prosthetics could improve and extend many lives, while renewable energy sources could help clean up the environment and lessen the deleterious health effects of air pol ution, a major and growing issue, particularly in developing economies. Many of these technologies could change how and what consumers buy, or alter overal consumption of certain resources such as energy and materials. Others could fundamental y change the nature of work for many employees around the world, both in manufacturing and knowledge work. Almost every technology on our list could change the game for businesses, creating entirely new products and services, as wel as shifting pools of value between producers or from producers to consumers. Some, like automation of knowledge work and the mobile Internet, could also change how companies and other organizations structure themselves, bringing new meaning to the anytime/ anywhere work style. With automation of knowledge work tasks, organizations that can augment the powers of skil ed workers stand to do wel .
Exhibit E6 How disruptive technologies could affect society, businesses, Primary Secondary Other potential impact and economies Implications for Implications for established businesses Implications for economies individuals and societies and other organizations and governments Changes Creates Creates Shifts Shifts quality of Changes opportu- new surplus surplus from Changes Drives Changes Poses new life, health, patterns of Changes nities for products between producers organi- economic comparative regulatory and envi- consump- nature of entre- and producers to zational growth or advantage Affects and legal ronment tion work preneurs services or industries consumers structures productivity for nations employment chal enges Mobile Internet Automation of knowledge work Internet of Things Cloud technology Advanced robotics Autonomous and near- autonomous vehicles Next- generation genomics Energy storage 3D printing Advanced materials Advanced oil and gas exploration and recovery Renewable energy SOURCE: McKinsey Global Institute analysis E6, also BN2
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 21 Each of these technologies has significant potential to drive economic growth and even change the sources of comparative advantages among nations. Energy technologies such as unconventional oil and gas and energy storage could power overal economic growth, while technologies such as advanced robotics and 3D printing could foster increased productivity and growth in the manufacturing sector. These types of impacts could help nations develop and exploit their unique resources and capabilities in new ways, potential y shifting the global center of gravity across sectors and regions. Many of these technologies pose new regulatory and legal chal enges. Some, such as autonomous vehicles, will require sensible regulatory regimes to help foster their growth and realize their benefits. Next-generation genomics and Internet of Things wil need appropriate controls to help avoid accidents or misuse. As these disruptive technologies continue to evolve and play out, it wil be up to business leaders, entrepreneurs, policy makers, and citizens to maximize their opportunities while dealing with the chal enges. Business leaders need to be on the winning side of these changes. They can do that by being the early adopters or innovators or by turning a disruptive threat into an opportunity. The first step is for leaders to invest in their own technology knowledge. Technology is no longer down the hal or simply a budget line; it is the enabler of virtual y any strategy, whether by providing the big data analytics that reveal ways to reach new customer groups, or the Internet of Things connections that enable a whole new profit center in after-sale support. Top leaders need to know what technologies can do and how to bend it to their strategic goals. Leaders cannot wait until technologies are ful y baked to think about how they wil work for—or against— them. And sometimes companies wil need to disrupt their own business models before a rival or a new competitor does it for them. One clear message: the nature of work is changing. Technologies such as advanced robots and knowledge work automation tools move companies further to a future of leaner, more productive operations, but also far more technological y advanced operations. The need for high-level technical skil s wil only grow, even on the assembly line. Companies wil need to find ways to get the workforce they need, by engaging with policy makers and their communities to shape secondary and tertiary education and by investing in talent development and training; the half-life of skil s is shrinking, and companies may need to get back into the training business to keep their corporate skil s fresh. The scope of impact of the technologies in this report makes clear that policy makers could benefit from an informed and comprehensive view of how they can help their economies benefit from new technologies. Policy makers can find ways to turn the disruptions into positive change; they can encourage development of the technologies that are most relevant to their economies. In many cases, such as in next-generation genomics or autonomous vehicles, the proper regulatory frameworks wil need to be in place before those technologies can blossom ful y. In other cases governments may need to be the standards setters or the funders of the research that helps move ideas from science labs into the economy. In still others, they wil need to draw the lines between progress and personal rights.
22 The chal enge for policy makers—and for citizens—is enormous. It is a good time for policy makers to review how they address technology issues and develop a systematic approach; technology stops for no one, and governments cannot afford to be passive or reactive. The time may be right, for example, to rethink how governments measure the economic impact of technology—to look beyond GDP and employment and look for metrics that truly capture the value added (or put at risk) when new technologies take hold.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 23 Disruptive technologies at a glance: Word cloud of report contents5 5 www.wordle.com; McKinsey Global Institute analysis.
24 Introduction: Understanding the impact of technologies The power of new technologies is everywhere. They change how businesses make money and how we live and work, sometimes with amazing speed. Social media was practical y unknown a decade ago, yet almost one bil ion people now have Facebook accounts; in fact, entirely new ways of socializing and interacting with friends, family, and col eagues have become the norm.6 Around the world, hundreds of mil ions of people have been lifted out of poverty as developing nations have adopted the technologies that drove growth in advanced economies in earlier times. Today, technologies such as the mobile Internet are helping to accelerate economic development, al owing mil ions of people in remote areas of developing regions to leapfrog into the 21st-century global economy. Technology’s power is particularly transformative in business. Technology can create immense value, but it often does so through a highly disruptive process. In the past, technological change has reordered industry after industry. Profit pools have shifted between owners of capital, labor, and consumers. Incumbent businesses have lost out, and startups have become dominant players. Whole product lines have been relegated to niches or forgotten altogether. Yet despite the increasingly notable presence of technology in our world, the ability to ful y measure its impact remains limited. We notice the effects of new technologies as they rapidly change our work routines, the way we spend our leisure time, or the products and services we use (often, increasingly, for free). We experience the benefits of new technologies in profound ways when they save or extend our lives or those of our loved ones. But existing economic statistics such as GDP struggle to ful y account for this value, which is often realized as consumer surplus and can take decades to show up in the numbers. Better approaches are needed to measure the ful economic impact of technologies, both to evaluate their potential and to set an appropriate course. HOW TECHNOLOGY DRIVES GROWTH Since the start of the Industrial Revolution more than 250 years ago, the global economy has been on a steep growth trajectory propel ed by a series of advances in technology (see Exhibit 1). From steam engines that replaced water mil s to electricity, telephones, automobiles, airplanes, transistors, computers, and the Internet, each new wave of technology has brought about surges in productivity and economic growth, enabling both efficient new methods for performing existing tasks and giving rise to entirely new types of businesses. Certain technologies, particularly general-purpose ones such as steam power or the Internet that can be applied across economies, have massive and disruptive effects. In this report we focus on the potential impact of such economical y disruptive technologies. 6 eMarketer, February 2012.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 25
Exhibit 1 Since the Industrial Revolution, the world has experienced an unprecedented rise in economic growth that has been fueled by innovation Estimated global GDP per capita $ 10,000 1,000 First Second Industrial Industrial Revolution Revolution 1760s to 1840s 1860s to 1920s 100 0 50 100 150 200 250 1550 1600 1650 1700 1750 1800 1850 1900 1950 2000 Technology Printing First Efficient Mass- Internal Internet advancements press steam steam produced combustion engine engine steel engine 1450 1698 1769 1855 1860 1970s Today SOURCE: Angus Maddison, “Statistics on World Population, GDP and Per Capita GDP, 1–2008 AD”; McKinsey Global Institute analysis Technological progress is not the only force that drives transformative growth in economies; for example, US growth during the 1970s was driven by the entry of mil ions of women and baby boomers into the labor force. However, technological advances have been an especial y valuable source of growth because they tend to be “non-rival” in nature, meaning they can be used over and over, benefiting different users and driving increasing returns. And unlike other sources of growth, such as increases in the labor force, the effects of technology do not go away.7 General-purpose technologies are particularly powerful. They are not only non-rival and long-lasting, but their pervasiveness also makes them especial y disruptive. The Internet is an excel ent example. It introduced new ways of communicating and using information that enabled major innovations, imposing new rules from outside on al sorts of industries, rearranging value chains and enabling new forms of competition. In industry after industry, Internet-enabled innovations brought transparency to pricing, disrupted commercial relationships, created new customer expectations, and made old business models obsolete. Napster and iTunes al but eliminated record stores, online booking systems have made travel agents largely redundant, and Amazon has forever changed both booksel ing and the book publishing industry as a whole. General-purpose technologies also tend to shift value to consumers, at least in the long run. This is because new technologies eventual y give al players an opportunity to raise productivity, driving increased competition that leads to lower prices. General-purpose technologies can also enable—or spawn—more technologies. For example, steam power enabled the locomotive and railroads, and the printing press accelerated learning and scientific discovery. General- 7 Economists cal technology-driven growth “intensive” (meaning a change in the rate inputs are converted to outputs), as opposed to extensive growth, which involves increasing inputs into the system.
26 purpose technologies can take many forms—including materials, media, and new sources of energy—but they al share the ability to bring about transformative change (see Box 4, “Three general-purpose technologies that changed the world”). Box 4. Three general-purpose technologies that changed the world The development of steel-manufacturing technology enabled the spread of a new material that accelerated growth and innovation in the Industrial Revolution. Steel is a stronger, lighter, and more ductile material than iron, but it was not easily produced until Henry Bessemer developed the Bessemer process that blew air through molten iron to remove the impurities. This technological advance enabled rapid, inexpensive mass production of steel, which could be done with mostly unskilled labor.1 Steel was quickly adopted for tools and machinery and in construction, ship-building, trains, and later in automobiles. Steel became an engine of growth helping double global GDP per capita between 1850 and 1900. Since then, steelmaking has been used repeatedly as an engine of growth for developing economies. The printing press, one of the first great information technologies, il ustrates how a general-purpose technology can have unpredictable effects. First used as a way to make the bible accessible, the printing press almost immediately became the agent of seismic social disruption in Europe as the leaders of the Reformation adopted the technology to print the tracts and pamphlets that spread the movement at unprecedented speed. Next, printing presses helped spark the scientific revolution— and the Enlightenment—by disseminating research and discoveries across the continent. Indirect effects included accelerated city growth; between 1500 and 1600, cities with printing presses grew 60 percent faster than other cities.2 Some historians attribute Europe’s rapid growth and global influence and the eclipse of Islamic nations after the 15th century to rapid adoption of printing in Europe and slow adoption in Islamic economies.3 The story of the electric dynamo (the first type of electric motor) demonstrates that unleashing the ful disruptive potential of new technology can be a long and difficult process. The electric dynamo represented a major improvement over existing steam- and water-powered engines because manufacturing stages or workstations no longer had to be tied to central power shafts in each factory. This al owed for improved factory organization and increased efficiency. However, as Stanford economist Paul David has noted, it took two decades (between 1900 and 1920) for this technology to reach 50 percent of factories in the United States and several more decades for the ful impact to be seen in productivity numbers.4 This was because firms were heavily invested in the legacy technology of the day (steam) and adopting the dynamo required redesigning equipment and reconfiguring facilities. David argues that similar factors could explain the observed lag between adoption of IT systems and measurable productivity increases. 1 Kathryn Kish Sklar, Florence Kel ey and the nation’s work: The rise of women’s political culture, 1830–1900, Yale University Press, 1995. 2 J. E. Dittmar, “Information technology and economic change: The impact of the printing press,” The Quarterly Journal of Economics, volume 126, number 3, 2011. 3 Metin Cosgel, The political economy of law and economic development in Islamic history, University of Connecticut Department of Economics working paper number 2012–44, December 2012, and Jared Rubin, “The printing press, reformation, and legitimization,” Stanford, Stanford University. 4 Paul David, “The dynamo and the computer: An historical perspective on the modern productivity paradox,” American Economic Review, volume 80, number 2, May 1990.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 27 Some economists today raise concerns that technology-driven growth could be slowing down. Economists Robert Gordon of Northwestern University and Tyler Cowen of George Mason University both cite slowing growth in productivity in advanced economies in recent years, arguing that technological advances such as the Internet may not have the same power to drive growth as prior generations of technology, such as those that occurred during the first and second Industrial Revolutions.8 Neither technology skeptics nor optimists can predict the future. Technologies and innovations are diffused and adopted at unpredictable rates. Whol y unanticipated applications may arise and become dominant, while the most obvious potential uses may not pan out. Moreover, when technologies are commercialized and widely used, the ways in which their impacts are measured can provide a distorted picture. Most metrics focus on industry impact—the amount of GDP generated by the production and consumption of a new technology in sectors where there is clear and direct impact (for example, how many microchips are made and then sold in computers). This misses the economic surplus that accrues to users, which can be the largest pool of value from disruptive technologies (such as the Internet). It also ignores effects on third parties, like children afflicted with asthma as a result of poor air quality or fishermen whose livelihoods are affected by water pol ution. The debate over how to measure technology impact is ongoing. In the 1980s economist Robert Solow caused a stir when he noted that despite the large investments that had been made in information technologies, there was no evidence of higher productivity in the service industries (e.g., banking) that had made the largest investments. GDP and other growth accounting metrics of IT impact do not ful y account for improved quality of outputs through use of technology. Nor do they measure the surplus that users capture through improvements in quality and other benefits that new technologies provide. In fact, GDP doesn’t directly measure any aspect of sustainability—whether in terms of environment, debt levels, or income distribution. Another problem with judging the economic impact of technologies is timing. Because technologies often create value in unpredictable ways, early assessments frequently turn out to be misleading. For example, many social media technologies have been dismissed as trivial based on how they are used by consumers. However, when applied to complex business organizations to improve communications, col aboration, and access to knowledge, the same tools that are used to share links to cute cat videos have enormous potential to improve the productivity of knowledge workers.9 Realizing the ful value of new technology can also be a long, difficult process. The electric dynamo was a major innovation that ushered in a revolution in manufacturing during the early 20th century but nevertheless took decades to reach widespread adoption and drive major productivity impact. Final y, the rate of adoption for a technology can vary a great deal from one economy to another. While it is true that productivity and GDP growth have been modest in the United States—which is on the leading edge of technology adoption—rapid technology adoption (albeit of older technologies) is driving 8 Economists describe this as a state of diminishing returns. 9 The social economy: Unlocking value and productivity through social technologies, McKinsey Global Institute, July 2012.
28 growth in developing economies. For example, mobile phone carrier Roshan has become Afghanistan’s largest employer by introducing the kind of mobile technology that is now several generations behind in advanced economies.10 Some economists have used alternative metrics to estimate the true impact of technologies. For example, Robert Fogel, a Nobel laureate at the University of Chicago, calculates social savings from technologies by estimating what it would cost society to accomplish a task in the same way that it did before an innovation was adopted (e.g., comparing the cost of transportation using railroads to the cost in a hypothetical scenario in which railroads were never adopted).11 Today, policy makers are increasingly aware of the limits of GDP. Institutions like the Organization for Economic Cooperation and Development, the European Commission, and the United Nations have al examined or adopted alternatives ranging from the Human Development Index to Bhutan’s Gross National Happiness measure. We are not technology cheerleaders—or pessimists. We believe, however, that there is reason for optimism. As we examine potential y economical y disruptive technologies on the horizon, we see significant potential for these technologies to raise productivity, disrupt existing business models, and create new profit pools. We also see that this growth wil be accompanied by risks and chal enges—as has always been the case for technology-led growth. As Erik Brynjolfsson and Andrew McAfee have argued, some advances that have the potential to drive productivity growth, such as advanced robotics and automated knowledge work, could also cause worrisome employment effects.12 As new technologies come into use, society wil need to continual y balance their benefits and risks. Consumers can be relied upon to embrace technologies that make their lives more convenient and provide new sources of entertainment. Businesses and public-sector institutions wil not forgo the productivity gains and other benefits that new technologies wil make possible. We also believe that over the long term and on an economy-wide basis, productivity growth and job creation can continue to grow in tandem, as they general y have historical y, if business leaders and policy makers can provide the necessary levels of innovation and education.13 However one measures its impact, the role of technology is growing in our economy and in society. The pace and direction of technological progress increasingly determines who gets hired, how our children are educated, how we find information and entertainment, and how we interact with the physical world. This puts the onus on society to find the most meaningful measures of the value derived from new technologies so that we can truly understand and control what is happening to our economies and our lives. In addition to GDP measures that 10 “Shining a light,” The Economist, March 8, 2007. 11 Tim Leunig, “Social Savings,” Journal of Economic Surveys, volume 24, issue 5, December 2010. Other attempts to supplement GDP measures include Amartya Sen’s Human Development Index (HDI), which we believe has limited applicability for estimating the potential economic impact of a particular technology. We have not used HDI or social savings in our estimates. 12 Erik Brynjolfsson and Andrew McAfee, Race against the machine: How the digital revolution is accelerating innovation, driving productivity, and irreversibly transforming employment and the economy, Digital Frontier Press, 2011. 13 Growth and renewal in the United States: Retooling America’s growth engine, McKinsey Global Institute, February 2011.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 29 focus on economic activity, we need metrics that account for true value—such as the consumer surplus that arises when a student using a tablet computer suddenly connects the dots and can solve a math equation on her own, or the value that an elderly person might place on the ability to move without assistance. Many efforts have been made to supplement GDP as a measure of value, but every alternative has its chal enges, and GDP continues to dominate global discourse and decision making. This report attempts to examine and estimate the potential economic impact of disruptive technologies using a consistent methodology that includes consumer surplus. But producing broad estimates of the potential economic impact of specific technologies is easy compared to the chal enges of measuring the actual, ful value that technologies create in the global economy. This chal enge is likely to remain—and grow—over the coming decade. As the saying goes, what gets measured is what gets done. As leaders think about realizing and capturing the ful value of economical y disruptive technologies, this idea might serve as a cal to action.
30 1. Mobile Internet The use of mobile Internet technology is already widespread, with more than 1.1 bil ion people currently using smartphones and tablets. The rapid and enthusiastic adoption of these devices has demonstrated that mobile Internet technology is far more than just another way to go online and browse. Equipped with Internet-enabled mobile computing devices and apps for almost any task, people increasingly go about their daily routines using new ways to understand, perceive, and interact with the world. In a remarkably short time, mobile Internet capability has become a feature in the lives of mil ions of people, who have developed a stronger attachment to their smartphones and tablets than to any previous computer technology.14 However, the ful potential of the mobile Internet is yet to be realized; over the coming decade, this technology could fuel significant transformation and disruption, not least from the possibility that the mobile Internet could bring two bil ion to three bil ion more people into the connected world and the global economy. Mobile computing devices and applications are evolving every day. New devices now incorporate features such as ultra–high resolution screens with precise touch sensing, graphic-processing power rivaling that of gaming consoles, and new kinds of sensors. The mobile Internet is also being offered in entirely new formats, such as wearable devices. New 4G wireless networks offer increasingly fast data speeds, al owing users to seamlessly transition from home broadband and office Wi-Fi to mobile voice and data services. New mobile software and apps offer a wide range of capabilities, effectively placing the capabilities of an array of gadgets (including PCs) in a mobile package that provides voice cal ing, Internet access, navigation, gaming, health monitoring, payment processing, and cloud access. We estimate that for the applications we have sized, the mobile Internet could generate annual economic impact of $3.7 tril ion to $10.8 tril ion global y by 2025. This value would come from three main sources: improved delivery of services, productivity increases in select work categories, and the value from Internet use for the new Internet users who are likely to be added in 2025, assuming that they wil use wireless access either al or part of the time. The prospect of three bil ion more consumers joining the digital economy could represent an unprecedented growth opportunity. Entrepreneurs in developing economies might be able to compete global y in online commerce, and global players wil have a new channel to reach the fastest-growing markets. Consumers in both advanced and developing economies stand to benefit substantial y from mobile Internet usage, as they have from the Internet itself—consumers have 14 A recent survey indicated that more people would prefer to leave home without their wal et than without their smartphone (“Consumer priorities: Choice of taking wal et or smartphone to work in 2012,” Statista, www.statista.com/statistics/241149/consumer-choice-of-taking- wal et-or-smartphone-to-work-in-2012/); the McKinsey iConsumer survey indicated that in the United States, 61 percent of smartphones are used only by their owner, versus 40 percent for laptops.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 31 captured roughly two-thirds of the economic surplus generated by the Internet.15 The mobile Internet also has great potential to improve delivery and raise productivity in health care, education, and other public and social services. DEFINITION We define the mobile Internet as a combination of mobile computing devices, high-speed wireless connectivity, and applications. Today, smartphones and tablets are the devices used to access the mobile Internet, but new forms are constantly emerging. In coming years, mobile Internet devices could wel be smal er, far more powerful, more intuitive, wearable, and packed with many types of sensors. With every new cycle of updates and models, tablets and smartphones are gaining capabilities. The processing power of the average smartphone has increased by about 25 percent per year over the past five years, and the latest processors can adeptly juggle multiple resource-intensive applications and produce vivid graphics. Smartphones and tablets are packed with sensors, including accelerometers and location sensors, and more recent models now include sensors that monitor temperature, humidity, and air pressure, as wel as infrared sensors and sensors that detect screen proximity, making phones easy to use in any light and extending battery life.16 Meanwhile, mobile Internet technologies are becoming more richly and intuitively interactive (see Box 5, “Vision of a connected world”). Apple’s Siri and Google Now both offer accurate voice recognition. Gesture recognition, already in wide use in gaming systems, is being adapted to mobile Internet devices; for example, the Samsung Galaxy S4 phone al ows users to browse by waving their hands in front of the screen.17 Wearable devices such as Google Glass and smart watches wil soon be available as wel . In addition to advances in devices, progress is being made in high-speed mobile connectivity. Today, mobile devices connect to the Internet via cel ular networks (3G and 4G/LTE networks) or Wi-Fi networks used in offices, factories, homes, and public spaces. Over the coming decade, network advances could include 5G cel ular networks (the as yet unspecified next-generation standard), satel ite services, and possibly long-range Wi-Fi. These technologies wil need to fight for increasingly scarce wireless frequencies, but new approaches to dynamical y sharing spectrum, such as software-defined radios, could help ease the crunch. At the same time, mobile operators are creating application programming interfaces (APIs) that al ow developers to control quality of service and bandwidth, potential y enabling new tiered pricing models.18 A crucial element in the success of mobile Internet use is the software applications (apps) that deliver innovative capabilities and services on devices. These apps help make mobile Internet use very different from using either conventional phones or computers, providing location-based services (directions 15 Internet matters: The Net’s sweeping impact on growth, jobs, and prosperity, McKinsey Global Institute, May 2011. 16 Stephanie Lanier, “Hidden features and sensors in the Samsung Galaxy S4,” AndroidGuys. com, April 10, 2013. 17 Brian X. Chen, “The Samsung Galaxy S4 review roundup,” Bits blog, The New York Times, April 24, 2013. 18 Venkat Atluri, Umit Cakmak, Richard Lee, and Shekhar Varanasi, Making smartphones bril iant: Ten trends, McKinsey & Company Telecom, Media, and High Tech Extranet, June 2012.
32 and location-based shopping tips, for example); personalized feeds of information and entertainment content; and constant online contact with friends, col eagues, and customers. Many of these apps seamlessly tap powerful resources on the cloud to fetch information or carry out tasks; when an iPhone owner asks a question of Siri, Apple’s voice-activated search interface/digital assistant (see Chapter 2, “Automation of knowledge work”), for example, the heavy processing takes place on the cloud (see Chapter 4, “Cloud technology”). Apps are crucial to the potential impact of mobile Internet use, multiplying its capabilities and potential y disrupting established business models (including older online business models in industries such as retail, banking, and media). By 2025, many apps could stil require some footprint on each device, but they could also offload much of their processing and storage to the cloud and might be accessed entirely over the Web using technologies such as HTML5, which can deliver applications via a Web browser without the need for downloading. Cloud enablement makes sense as devices become smal er and more seamlessly connected. Box 5. Vision of a connected world What wil the world be like when everyone is connected almost al the time? Imagine your morning routine. You strap on your smart watch, put on your smart glasses, and head to work. As you glance at the shoes arrayed in a department store window, information about them pops up in your line of vision. Descriptions of the shoes, along with prices, sizes, colors, and availability appear. You’re about to make a selection when an audio alarm sounds in your earpiece. The next express bus wil be at the corner in two minutes. You get to the bus stop just in time. The doors open and a display instal ed where the fare box used to be greets you by name and tel s you how much credit you wil have on your virtual transit card (stored on your smartphone) after this trip. Or imagine you are a farmer in a developing economy. Until recently, you have been limited to subsistence farming, producing only enough to feed your family. But since acquiring a smartphone, your fortunes have improved. You have participated in online instruction regarding irrigation and other farming techniques, which have raised your crop yield, and you now sell most of what you grow. Using your new Internet connection, you have linked up with other farmers to form a cooperative to buy seed, fertilizer, and equipment and to pool your crops at harvest time to get better prices. While your crops are growing, you use your phone to photograph them and use a smart app to automatical y grade them. These data are entered into the co-op’s database, which then gives you an estimated price. Every day you get automatic reports on world prices for your crops. When the harvest is sold, your phone alerts you when payment hits your bank account. Without leaving your vil age, you have joined the global economy.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 33 POTENTIAL FOR ACCELERATION The largest opportunity for growth of mobile Internet use and impact wil be in developing economies, where access to the Internet has important implications for economic development, potential y helping hundreds of mil ions of people to leapfrog into the 21st-century global economy. At this time, half of the world’s adult population uses no banking services. UNICEF estimates that more than 100 mil ion children do not attend school, and few of the farmers on the world’s 500 mil ion small-scale farms have broader access to markets.19 Mobile access to the Internet could address many of the basics needs of the developing world and enable its citizens to become participants in the global digital economy, including as entrepreneurs. In the developed world, growth wil continue to be driven by new devices and applications, as wel as by growing intensity of use as consumers and businesses come to rely on mobile Internet access in more aspects of their daily routines and operations. Mobile Internet technology consists largely of smartphones and tablets. Sales growth of these devices has been extremely rapid and wil likely continue at a brisk pace and perhaps even accelerate, particularly in developing economies.20 The number of smartphones in use grew 50 percent in 2012, and smartphones now account for 30 percent of mobile devices in use global y.21 Sales of smartphones are projected to reach 1.3 bil ion units per year in 2013, while tablet sales are expected to reach 200 mil ion units.22 Sales of smartphones and tablets exceeded personal computer sales in 2010, and in 2013 the number of smartphones and tablets in use is expected to exceed the instal ed base of personal computers.23 By 2025, nearly 80 percent of al Internet connections could be through mobile devices, and a majority of new Internet users could be using mobile devices as their primary or sole means of connecting to the Internet. As consumers spend more time online, the number and quality of Internet- based services are increasing, further driving demand. App downloads grew 150 percent in 2012, and an array of new mobile services have emerged. So- cal ed near-field payments (which use unpowered radio frequency chips to easily exchange data between devices) grew 400 percent in 2012, and are expected to increase 20-fold by 2016.24 These are the systems that al ow consumers to wave a phone near a point of sale terminal to make a payment, for example. Media and entertainment consumption on mobile devices has grown and is rapidly shifting viewers from cable and broadcast channels. Time spent playing video games, emailing, and text messaging on mobile phones grew 200 percent in the past four years. In the United States, an estimated 30 percent of al Web browsing and 40 percent of social media usage is now done on mobile devices. Time spent online on mobile phones is increasing at 25 percent per year, and data traffic on mobile devices has reached 15 percent of total Internet traffic.25 The rapid shift of activities and content consumption to mobile Internet devices could represent 19 Agriculture at a crossroads, International Assessment of Agricultural Knowledge, Science and Technology, Center for Resource Economics, 2009. 20 Yankee Group Global Mobile Forecast. 21 Ibid. 22 Ibid. 23 Mary Meeker, Internet trends, Kleiner Perkins Caufield & Byers, March 2013. 24 Yankee Group Global Mobile Forecast. 25 2012 McKinsey US iConsumer Survey; StatCounter Global Stats.
34 only the beginning of a long-term trend. It is possible that by 2025, a far higher percentage of media consumption could be dominated by this technology.26 The level of demand for mobile Internet access in developing economies will depend largely on how wel device makers and mobile Internet services tailor their offerings to the needs of people who are just entering the global consumer class. While Internet use has been growing by 25 percent a year in developing economies (compared with 5 percent in advanced economies), 64 percent of the population in developing economies are not yet connected.27 By 2016 developing economy markets are expected to be the largest source of smartphone market growth. India’s share of global smartphone sales is expected to grow from 2 percent in 2012 to 9 percent in 2016, and Brazil’s share is expected to rise from 2 percent to 5 percent.28 Continued reductions in the prices of smartphones and data plans should help sustain rapid adoption rates. Anticipated decomponent costs are expected to continue to decline, which could reduce producer costs for midrange smartphones by about 30 percent by 2016. The evolution of smartphone hardware and the emergence of other mobile devices should also help fuel sales and inspire new uses. Wearable devices such as the Google Glass not only make it possible to deliver al sorts of content in novel ways (for example, projecting images from the Web that appear to float in space in front of the wearer), but also make it possible to develop “augmented reality” applications that let the wearer step into virtual spaces just as virtual reality goggles do. This technology has obvious applications for entertainment and gaming, but these devices could also be used to help people in new ways. For example, a wearable mobile device could be programmed to help Alzheimer’s patients recognize people and remind them of what various objects around the home are. Instant translation apps on wearable devices could be used to read signs and menus, making travel to foreign lands far easier. Final y, as mobile Internet devices become more integrated into day-to-day life—and gain new capabilities—they have the potential to become intel igent personal assistants, capable of managing our schedules, answering questions, and even alerting us to important information (see Chapter 2, “Automation of knowledge work”). POTENTIAL ECONOMIC IMPACT In the uses we analyzed, mobile Internet usage could generate global economic impact of $3.7 trillion to $10.8 trillion per year by 2025 (Exhibit 2). Half of this potential value could come from using mobile devices to spread Internet access in developing regions. More than 3.5 bil ion citizens in developing economies are expected to have access to the Internet in 2025, more than two bil ion of them via mobile Internet services. People who have had poor access to education, health care, and government services and have never participated in the formal economy could become participants in the global economy through the Internet. 26 2012 McKinsey US iConsumer Survey. 27 Online and upcoming: The Internet’s impact on aspiring countries, McKinsey Global Institute, January 2012. 28 Worldwide quarterly mobile phone forecast, IDC, March 2012.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 35
Exhibit 2 Sized applications of mobile Internet could have direct economic impact of $3.7 trillion to $10.8 trillion per year in 2025 Potential economic impact of sized Potential applications in 2025 Estimated scope Estimated potential productivity or Sized applications $ tril ion, annual y in 2025 reach in 2025 value gains in 2025 $15.5 tril ion cost 70–80% mobile 10–20% cost of treating penetration among reduction in chronic chronic diseases patients accounting for disease treatment 95% of health-care through remote spending health monitoring $11 tril ion global K–12 adoption of 5–15% rise in spending on online/hybrid learning secondary education – Developed world: graduation rates 80–90%1 10–30% 0.9– – Developing: 65–80% productivity gain in Health care 90–100% adoption in post-secondary, 2.1 post-secondary, corporate, and corporate, and government government education education 0.3– Education $0.9–1.2 tril ion Adoption by 90–100% of 60–75% cost 1.0
government governments for online savings on spending on or mobile services administrative tasks
customer-facing driven by labor
services efficiency Service Public 0.2– delivery sector 0.5 $7.2 tril ion cost 30–50% of retail 6–15% productivity of retail consumption gain of online Mobile devices used in hybrid retail versus 50% of purchases traditional 0.1– Retail 0.4 $3 tril ion in Implementation of 50% productivity global advanced electronic gain in managing transaction payments systems in transactions across revenue – 80–100% of advanced al stakeholders 0.2– economies Payments 0.3 – 65–80% of developing economies1 $19 tril ion in 80–90% of workers in 4–5% increase in Interaction 0.9– interaction advanced economies efficiency through workers 1.3 worker salaries 65–80% of workers in social technology Other developing economies via mobile worker produc- $15 tril ion in 80–90% of workers in 10–30% tivity2 Transaction 0.1– transaction advanced economies productivity gain workers 0.4 worker salaries 65–80% of workers in from time saved developing economies accessing Mobile devices needed information for 10% of work tasks Additional 1.0– consumer 4.8 3.6–4.9 bil ion 100% of users $500–1,500 per surplus mobile users user in developed world $300–1,000 per Other user in developing potential world applications (not sized) Sum of sized potential 3.7– economic 10.8 impacts 1 Estimates of adoption are based on Internet penetration rates in advanced and developing economies. 2 Estimates of potential economic impact for worker productivity applications exclude labor productivity impact sized as part of service delivery applications. NOTE: Estimates of potential economic impact are for some applications only and are not comprehensive estimates of total potential impact. Estimates include consumer surplus and cannot be related to potential company revenue, market size, or GDP impact. We do not size possible surplus shifts among companies and industries, or between companies and consumers. These estimates are not risk- or probability-adjusted. Numbers may not sum due to rounding. SOURCE: McKinsey Global Institute analysis 2
36 To estimate the potential economic impact of mobile Internet technology, we looked at how mobile applications could improve service delivery, raise productivity, and create value for consumers in time savings and convenience surplus.29 We have attempted to exclude the value of Internet use via fixed connections while including the value of the additional Internet use and value that users of fixed Internet connections derive from also using the mobile Internet. For potential mobile Internet users in 2025 who previously lacked Internet access (that is, new users in developing economies), we have treated al Internet use as incremental. Among the types of services that stand to benefit from mobile Internet technology, health care is one of the most promising. In just one application— management of chronic disease—this technology potential y could cut more than $2 tril ion a year in the projected cost of care by 2025. Today, treating chronic diseases accounts for about 60 percent of global health-care spending, and it could be more than $15 tril ion global y by 2025.30 Patients with conditions such as heart disease and diabetes could be monitored through ingestible or attached sensors, which can transmit readings and alert the patient, nurses, and physicians when vital signs indicate an impending problem, thus avoiding crises and the costs of emergency room visits or hospitalization. For example, the US Veterans Health Administration has provided remote monitoring devices to more than 70,000 patients with chronic diseases, combined with access to video chats with physicians. These patients used 20 to 50 fewer service resources than those in the control group.31 Taking into account hurdles such as patient resistance, the cost of chronic disease treatment could be reduced by 10 to 20 percent through better disease management via the use of mobile Internet access, a relatively conservative estimate. Such a reduction could drive a potential economic impact of $900 bil ion to $2.1 tril ion per year by 2025. In education, mobile computing has the potential to raise productivity and improve learning both inside and outside classrooms. In K–12 education, early experiments show promise for hybrid online/offline teaching models using tablets to increase lesson quality, improve student performance, and increase graduation rates. Based on studies of the effectiveness of hybrid teaching models that incorporate mobile devices in instruction, dril s, and testing (alongside traditional classroom teaching), an improvement in graduation rates of 5 to 15 percent could be possible. This assumes a gradual adoption rate, with most of the benefit coming closer to 2025, when more students wil have benefited from online learning via tablets for most of their K–12 years. In higher education, as wel as government and corporate training, such hybrid models could improve productivity by 10 to 30 percent. Over the next decade, most types of education and training could adopt Internet-based hybrid education, affecting bil ions of 29 We used recent case examples as a conservative means of estimating potential productivity gains from mobile Internet technology. Our estimates do not include uses that have yet to be adopted, although the history of this technology indicates that such inventions are likely. 30 Based on spending on chronic diseases in France, Canada, the United Kingdom, and the United States. 31 Based on a case study by the US Veterans Health Administration regarding chronic heart failure, diabetes, and chronic obstructive pulmonary disease, including more than 70,000 patients. See Andrew Broderick and David Lindeman, “Scaling telehealth programs: Lessons from early adopters,” Case Studies in Telehealth Adoption, The Commonwealth Fund, January 2013.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 37 individuals. The share delivered via mobile devices could have economic impact of $300 mil ion to $1.0 tril ion annual y. In the public sector, many citizen services (such as information requests, license applications, and tax payments), could become online services through mobile apps. Based on McKinsey experience, it is possible to raise productivity by 60 to 70 percent by moving functions such as tax refund services and vehicle registration renewals to online channels. Assuming high adoption rates motivated by the need to control government spending, the potential economic impact of delivering government services using mobile Internet technology could reach $200 bil ion to $500 bil ion per year by 2025. In the retail sector, mobile Internet usage has great potential to extend the reach of online and hybrid online shopping( for instance, visiting showrooms and then purchasing online). Based on differences in prices and margins between traditional retail and online stores, the productivity gain of delivering retail goods through a digital channel could be 6 to 15 percent, based on reduced labor, inventory, and real estate costs. By 2025, 30 to 50 percent of retail transactions (40 to 70 percent in advanced economies and 20 to 30 percent in developing economies) might take place online, with a potential economic impact of $100 bil on to $400 bil ion per year. For government and private-sector organizations, mobile payments represent a very large opportunity made possible by mobile Internet technology. Today, 90 percent of the more than three tril ion transactions made every year global y are stil cash transactions, and McKinsey analyses show that an electronic transaction can save 50 to 70 percent of processing costs over a paper transaction. In an advanced economy, moving to an increased share of electronic transactions could have a productivity benefit equivalent to 0.35 percent of GDP. The total potential economic impact of moving transactions to an electronic format is estimated to $200 bil ion to $300 bil ion per year in 2025. Mobile applications could have considerable impact on improving internal operations, from frontline workers to sales reps to highly paid knowledge workers. Frontline workers, for example, could use mobile Internet devices to manage equipment and physical assets more effectively, monitor supply chains, maintain the condition of vital equipment, and provide post-sale services (see Chapter 3, “The Internet of Things”). Boeing and BMW have developed virtual reality glasses for assemblers and mechanics that display online manuals and instructions explaining, for example, exactly how parts should fit together.32 A University of Chicago study recently found that giving iPads to medical residents reduced the time it took to schedule procedures and improved the residents’ ability to explain complicated diagnoses to patients using visual aids.33 For transaction workers such as sales reps, mobile devices are already showing potential to increase productivity by making pricing, options, configurations, financing terms, and other information instantly available. In business-to-business sales, this has increased close rates by 35 to 65 percent in some cases. 32 Clay Dil ow, “BMW augmented reality glasses help average Joes make repairs,” PopSci.com, September 3, 2009. 33 Mary Modahl, Tablets set to change medical practice, QuantiaMD report, June 15, 2011.
38 One of the biggest opportunities for operational improvements involves using mobile Internet technology to increase the productivity of knowledge workers, including so-cal ed interaction workers, a category that includes professionals, administrative support staff, and others whose jobs require person-to-person interaction and independent judgment. Such workers stand to benefit most from the use of social technologies that enable communications and col aboration, which could raise interaction worker productivity by 20 to 25 percent, particularly by reducing the time it takes to handle email, search for information, and col aborate with col eagues.34 Assuming that interaction workers spend about 25 percent of the work week away from their desks, mobile Internet access could improve their productivity by 4 to 5 percent. For transaction workers, we estimate a time savings of 1-2 hours per day; however, only 10 percent of worker time online, depending on mobile connection. The estimated potential economic impact of worker productivity gains achieved via the use of mobile Internet applications in internal operations could be $1.0 to $1.7 tril ion annual y by 2025. Consumer surplus wil make up a large portion of the potential impact of mobile Internet access. We base our estimate of value on survey research by the Interactive Advertising Bureau (IAB) Europe and McKinsey & Company that established how consumers in developed and developing economies value Internet services.35 The applications we considered for an analysis of the surplus that would accrue through mobile Internet use are email, social networks, entertainment (music, videos, and games), and Web services such as search and mapping. To gauge the value of Internet use in 2025, we estimate that usage will grow by between 6 and 9 percent annual y. Applying this estimate of consumer surplus to new users (two to three bil ion), as wel as the incremental value for existing users (and subtracting estimated fixed Internet use), the impact could be between $1.0 and $4.8 tril ion annual y by 2025. BARRIERS AND ENABLERS Reaching the ful potential of mobile Internet use wil require device makers to pack more computing power, sharper displays, multiple sensors, and powerful antennas into ever-smal er mobile devices. Adding to the complexity of the task, these features, processors, and data-intensive uses al raise power consumption—yet battery size cannot increase due to the size constraints of mobile devices. Since 2000, battery capacity has only doubled, while processing speed has increased roughly 12-fold.36 Progress is possible: there could be rapid advances in lithium batteries, while advanced nanomaterials such as graphene could be used in electrode coatings that improve battery performance (see Chapter 8, “Energy storage,” and Chapter 10, “Advanced materials”). Progress in battery technology wil be needed in order to bring mobile Internet access to places in the developing world that lack a reliable supply of electricity. If these performance improvements are not achieved, the ful economic potential of mobile Internet access might not be realized. As mobile Internet devices and connections continue to proliferate and the amount of data per user grows, availability of sufficient radio spectrum is becoming a growing concern. Over the past several years, as users have started 34 The social economy, McKinsey Global Institute, July 2012. 35 Internet matters, McKinsey Global Institute, May 2011 36 Venkat Atluri, et al., Making smartphones bril iant, McKinsey & Company Telecom, Media, and High Tech Extranet, June 2012.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 39 streaming more “rich media” (including TV shows), downloading apps, and using more complex websites, data traffic on mobile networks has grown by 80 percent to 100 percent annual y. The growing scarcity of unused wireless spectrum has prompted some telecom carriers to spend bil ions of dol ars buying spectrum rights in public airwave auctions or from other companies.37 This shortage is creating an urgent need for better use of frequencies, such as dynamical y al ocating spectrum among users. Creating the infrastructure to increase global high-speed Internet coverage wil require significant capital expenditure. The expected investment in infrastructure necessary to facilitate mobile Internet usage could be as high as $300 bil ion annually in 2025, including cost for installation (the majority of which would be in Asia). The economic potential of mobile Internet usage may not be ful y realized if sufficient wireless spectrum capacity cannot be made available. IMPLICATIONS The economic impact of mobile Internet usage is potential y massive, and its effects could be highly disruptive across a wide range of sectors. Consumers, business leaders, and policy makers al have a stake in seeing mobile Internet usage spread and take on greater capabilities. And al of these stakeholders wil also have to grapple with chal enges that could limit the realization of this ful potential. Makers of mobile Internet devices, software providers, and other technology suppliers are likely to compete intensely to develop new and better products. They wil need to adopt new components, explore new product forms (such as “wearable” devices), and keep up with constantly evolving consumer preferences. Wireless carriers could face increasing chal enges in profiting from the growth of mobile Internet use. Intensifying competition between wireless carriers is already squeezing margins on mobile data plans. As more people get mobile Internet access and new data-intensive uses—such as streaming video programming— become more widely used, networks could slow down. Wireless carriers wil need to address these network capacity constraints, balancing capital investment with long-term profitability. They wil also face difficult decisions regarding whether to upgrade existing infrastructure or leapfrog to more advanced platforms through expansion or acquisition. For incumbent Internet businesses, growing mobile Internet use poses multiple chal enges. To remain competitive, they must adapt their services to mobile networks, often requiring significant investment. Today, the biggest chal enge for many online businesses is capturing revenue as Internet traffic goes mobile. At the same time, low barriers to launching a mobile-based online business make it easier for upstarts to chal enge established online players. Mobile Internet use is attracting entrepreneurs and capital, and the next innovators could come from literal y anywhere: more than 143,000 Internet-related businesses are started in developing economies every year.38 37 One recent example of a wireless spectrum purchase is Verizon’s purchase of spectrum for $3.9 bil ion from Time Warner Cable, Comcast, Cox Communications, and Bright House Networks. 38 Online and upcoming: The Internet’s impact on aspiring countries, McKinsey Global Institute, January 2012.
40 For many businesses, the prospect of three bil ion more consumers coming into the digital economy could represent an unprecedented growth opportunity, but one that wil require fresh thinking and new approaches. Products and services must not only fit within more limited budgets, but must also cater to very different tastes and priorities. Business leaders wil have to learn how to please these new consumers and effectively meet their needs. Business leaders wil also need to identify employee functions that could be performed more efficiently or more effectively on a mobile platform. They wil need to consider how performance could be enhanced by enabling increased mobility, augmenting worker knowledge and capabilities or facilitating col aboration and social interaction. Consumers in both advanced and developing economies stand to benefit substantial y from mobile Internet usage, as they have from the Internet itself. Consumers have captured roughly two-thirds of the economic surplus generated by the Internet, and this pattern could hold true for mobile Internet use as wel . The proliferation of mobile devices has also raised important societal concerns, however. This includes the effect of excessive screen time on child development and potential loss of productivity as a result of a distracted workforce. Entrepreneurs and companies that can develop products and services that address and al eviate these issues could create significant value. Policy makers around the world must learn how to use mobile Internet access to improve services, increase productivity, and drive economic development. Governments can also play a crucial role in accelerating the adoption of mobile Internet access by funding basic research and helping to overcome major barriers, for example by allocating scarce spectrum. In developing economies, governments have much to gain from mobile Internet usage as a driver of development and employment. As it has done in other places, the shift of business to the Internet can be disruptive to employment. However, for every job that is lost due to the Internet in smal and medium- size enterprises survey data indicates that in some developing countries 3.2 new jobs could be created.39 Furthermore, mobile Internet technologies could provide access to education and have a direct impact on critical issues such as malnutrition and malaria by helping to spread knowledge and distribute supplies where they are needed most. Even if the ultimate economic value of mobile Internet technology fal s far short of its potential, mobile Internet use wil almost certainly have lasting and profound effects. In a few years, smartphones and tablets have had enormous impact on business sectors ranging from personal computers to TV networks. They have opened up new paths to economic inclusion and growth in developing economies. But perhaps the most enduring effects wil come from our now ubiquitous connectivity, which in many societies has changed human behavior profoundly. Mobile Internet access does not endow users with superhuman powers, but it has shown the potential to make our lives more convenient, effortless, and enjoyable. 39 Online and upcoming: The Internet’s impact on aspiring countries, McKinsey Global Institute, January 2012.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 41 2. Automation of knowledge work A confluence of advances in computational speed, machine learning, and natural user interfaces has brought computing to an important milestone: computers are now becoming capable of doing jobs that it was assumed only humans could perform.40 Computers, for example, can now act on “unstructured” commands— answering a question posed in plain language—and even make subtle judgments. They can sift through massive amounts of information to discern patterns and relationships. They can “learn” rules and concepts based on examples or simply by crunching data. And, with advanced interfaces and artificial intel igence software, they can understand and interpret human speech, actions, and even intentions from ambiguous commands. In short, computers can increasingly do many of the tasks that are currently performed by knowledge workers (see Box 6, “The vision: The power of omniscience”). The commercialization of computers with this level of intel igence over the coming decade could have massive implications for how knowledge work is conducted. Such tools could both extend the powers of human workers and al ow them to offload tedious detail work. But these advanced tools could also ultimately lead to some jobs being automated entirely. Automation has already swept through manufacturing and transaction work (tasks that consist of executing simple exchanges, such as taking deposits or checking customers out of a grocery store). When it comes to knowledge work, the impact of automated tools could be less direct. Knowledge work jobs general y consist of a range of tasks, so automating one activity may not make an entire position unnecessary (the way welding robots make welders redundant, for example). In addition, knowledge work has become more complex, in large part due to information technology, creating demand for workers with new skil s who can perform new kinds of tasks. Nonetheless, there is potential for emerging tools to have a dramatic economic impact by 2025. In the applications we sized, we estimate that knowledge work automaton tools and systems could take on tasks that would be equal to the output of 110 mil ion to 140 mil ion ful -time equivalents (FTEs). It is possible that this incremental productivity—which does not include any estimate of the value of higher quality output due to better knowledge tools—could have as much as $5.2 tril ion to $6.7 tril ion in economic impact annually by 2025.41 As with advanced robotics (see Chapter 5), automation of knowledge work could bring great societal benefits—such as improved quality of health care and faster drug discovery—but may also spark complex societal chal enges, particularly in employment and the education and retraining of workers. This technology could 40 Erik Brynjolfsson and Andrew McAfee, Race against the machine: How the digital revolution is accelerating innovation, driving productivity, and irreversibly transforming employment and the economy, Digital Frontier Press, 2011. 41 Cost of employment numbers include salaries and benefits.
42 change the nature of work for many people, requiring innovation to ful y realize its potential while managing its risks. Box 6. The vision: The power of omniscience It’s 2025 and you arrive at your desk for another day at work. As you take your seat, the day’s appointments are diplayed in front of you and your digital assistant begins to speak, giving you a quick rundown of the 43 new posts on the departmental communications site. Three are important for today’s meetings; the rest wil be summarized by the system and sent in the daily report. The assistant notes that al the reports and multimedia presentations have been uploaded for your meetings. Now it’s time for the tough part of the day: your doctor appointment. You received a request for an appointment yesterday when your biosensor alerted your digital physician to a change in your blood pressure. Your vital signs are scanned remotely, and the system cross-checks this information with journal cases, your family’s history of hypertension, your diet and exercise routines, and the vital signs of other men your age. Good news: “You don’t need drugs, but you do need to stop eating fast food and skipping the gym,” your computerized doctor says. Relieved, you stop at the gym on the way home and ask your mobile device to order a salad to be delivered when you get home. DEFINITION We define knowledge work automation as the use of computers to perform tasks that rely on complex analyses, subtle judgments, and creative problem solving. Knowledge work automation is made possible by advances in three areas: computing technology (including processor speeds and memory capacity), machine learning, and natural user interfaces such as speech recognition technology. These capabilities not only extend computing into new realms (for example, the ability to “learn” and make basic judgments), but also create new relationships between knowledge workers and machines. It is increasingly possible to interact with a machine the way one would with a coworker. So, instead of assigning a team member to pul al the information on the performance of a certain product in a specific market or waiting for such a request to be turned into a job for the IT department, a manager or executive could simply ask a computer to provide the information. This has the potential to provide more timely access to information and raise the quality and pace of decision making and, consequently, performance.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 43 Advances in software, especial y machine learning techniques such as deep learning and neural networks, are key enablers of knowledge work automation.42 These techniques give computers the ability to draw conclusions from patterns they discern within massive data sets (anything from al legal cases of the past 20 years to data concerning the way in which molecular compounds react with one another). Importantly, computers with machine learning capabilities no longer rely only on fixed algorithms and rules provided by programmers. They can also modify and adjust their own algorithms based on analyses of the data, enabling them to “see” relationships or links that a human might overlook. Moreover, these machines can “learn” more and get smarter as they go along; the more they process big data, the more refined their algorithms become. Final y, advances in user interfaces, such as speech and gesture recognition technology, give computers the ability to respond directly to human commands and requests. So, for example, managers wil no longer have to learn computer syntax or send a request to the IT department to get a question answered by a computer. Apple’s Siri and Google Now use such natural user interfaces to recognize spoken words, interpret their meanings, and act on those meanings. This requires significant processing power and sophisticated software to filter words from background noise, to parse sentences, and to then make “smart” guesses about the intent of the query—for example, by offering up the names of sushi restaurants when the speaker asks for Japanese food. POTENTIAL FOR ACCELERATION Using these emerging tools, it wil be possible to automate an expanding variety of knowledge worker tasks. As noted, the force of automation has already swept through manufacturing and transaction work, with profound impact. To put this in perspective, in 40 years of automating transaction work, in some US transaction occupations, more than half of the jobs were eliminated. ATMs took on the work of bank tel ers, self-serve airline reservation systems replaced travel agents, and typists al but disappeared (Exhibit 3). Today, total global employment costs are $33 tril ion a year and, on current trend, could reach $41 tril ion by 2025. We analyzed a subset of knowledge worker occupations with employment costs that could reach $14 tril ion by 2025. These workers—professionals, managers, engineers, scientists, teachers, analysts, and administrative support staff—represent some of the most expensive forms of labor and perform the most valuable work in many organizations. Few of these workers have benefited from tools that can augment core aspects of their work involving decision making and judgment. 42 Many current machine learning approaches mimic aspects of the human brain. Neural networks simulate brain structures via interconnected layers of “artificial neurons,” which adaptively strengthen or weaken their interconnections based on experience. Deep learning technology makes use of algorithms that form a learning hierarchy in which higher-level concepts are defined using layers of lower-level concepts (often using neural networks). Some machine-learning algorithms don’t require labeling or preclassification of their training examples and can instead identify their own categories and concepts (e.g., by cluster analysis).
44 Rapid advances in underlying technologies are reducing costs and boosting performance, making knowledge automation more attractive. Computing power continues to grow exponential y (approximately doubling every two years on a price/performance basis) and today a $400 iPhone 4 offers roughly equal performance (in mil ions of floating point operations per second or MFLOPS) to the CDC 7600 supercomputer, which was the fastest supercomputer in 1975 and cost $5 mil ion at the time.43 These advances in computational power have been accompanied by significant strides in data storage systems, big data (the ability to process and analyze huge amounts of data, such as real-time location feeds from mil ions of cel phones), and cloud computing (which makes the computational power needed for knowledge work automation accessible to even individuals and smal businesses via Internet-enabled devices).
Exhibit 3 The number of transaction workers in the United States across some major job types declined more than 50 percent between 1970 and 2010 Decline in transactional jobs between 1970 and 20101 % workforce share decline for select highly automatable jobs Index: 100 = 1972 120 100 80 General clerks 60 Bookkeeping jobs 40 Secretaries Jobs 20 Typists almost Telephone automated operators 0 away 1972 1975 1980 1985 1990 1995 2000 2005 2010 1 Job types that can be scripted, routinized, automated (e.g., cashiers, receptionists, stock traders). Data are for the US private economy. Occupation data normalized in 1983 and 2003 to account for classification differences. SOURCE: US Bureau of Labor Statistics 1972–2010; McKinsey Global Institute analysis POTENTIAL ECONOMIC IMPACT Overal , we estimate the potential economic impact of knowledge automation tools in the types of work we assessed could reach $5.2 tril ion to $6.7 tril ion per year by 2025 due to greater output per knowledge worker (Exhibit 4). Of this total, the lion’s share ($4.3 tril ion to $5.6 tril ion) could be generated in advanced economies where wage rates are higher. In advanced economies, we estimate annual knowledge worker wages at about $60,000, compared with about $25,000 in developing economies, and project that increased automation could drive additional productivity equivalent to the output of 75 mil ion to 90 mil ion full- time workers in advanced economies and 35 mil ion to 50 mil ion ful -time workers in developing countries. 43 John V, “Apple iPhone 4S benchmark tests,” phonearena.com, October 20, 2011; Gordon Bel , “A Seymour Cray perspective,” Microsoft Research, November 1997.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 45
Exhibit 4 Sized applications of automation of knowledge work could have direct economic impact of $5.2 trillion to $6.7 trillion per year in 2025 Potential economic Potential impact of sized Estimated productivity Sized knowledge worker occupations in 2025 Estimated scope potential reach or value gains occupations $ tril ion, annual y in 2025 in 2025 in 2025 $4.4 tril ion in 50–65 mil ion $35,000 value 1.1– knowledge worker ful -time Clerical per FTE of 1.3 Common costs equivalents additional business 125 mil ion (FTEs) of work productivity functions Customer service 0.6– knowledge workers potential y and sales 0.9 automatable 0.8– Education $2.8 tril ion in know- 20–30 mil ion $50,000 value Social 1.0 ledge worker costs FTEs of work per FTE of sector
55 mil ion knowledge potential y additional services 0.3–
Health care workers automatable productivity 0.4
Science and 0.6– $2.2 tril ion in know- 15 mil ion $60,000 value engineering 0.7 ledge worker costs FTEs of work per FTE of Technical 35 mil ion knowledge potential y additional professions 0.4– workers automatable productivity IT 0.5 $2.9 tril ion in know- 15–20 mil ion $60,000 value 0.8– ledge worker costs FTEs of work per FTE of Managers 1.1 50 mil ion knowledge potential y additional workers automatable productivity 0.4– Finance 0.5 $1.5 tril ion in know- 10 mil ion $65,000 value Professional ledge worker costs FTEs of work per FTE of services 0.2– 25 mil ion knowledge potential y additional Legal 0.3 workers automatable productivity Other potential applications (not sized ) Sum of sized potential 5.2– economic impacts 6.7 NOTE: Estimates of potential economic impact are for some applications only and are not comprehensive estimates of total potential impact. Estimates include consumer surplus and cannot be related to potential company revenue, market size, or GDP impact. We do not size possible surplus shifts among companies and industries, or between companies and consumers. These estimates are not risk- or probability-adjusted. Numbers may not sum due to rounding. SOURCE: McKinsey Global Institute analysis To gauge the impact of automation across knowledge work, we looked at the number of employees and employment costs for 20 knowledge worker occupations across 11 countries (seven developed and four developing), which employ 75 percent of al workers in these occupations and whose employers pay about 80 percent of global employment costs. From this, we estimated the potential productivity improvement per unit of labor by comparing the estimated costs of relevant technology against the total cost of employment for potential y affected occupations. We found the largest potential impact in common business functions such as clerical and administrative work ($1.7 tril ion to $2.2 tril ion), fol owed by jobs in the social services sector such as education and health care ($1.1 tril ion to $1.4 tril ion), and then technical professions and management (about $1 tril ion each). Common business functions Many common business functions (for example, cal center sales, administrative support, and customer service) involve answering questions or carrying out tasks for other workers or customers. Adva 4 nces i n natural user interfaces (including software that can understand and act on questions using ordinary speech, rather than in the strict format and syntax of computer languages) could make many of these tasks automatable. It is possible that by 2025, productivity gains
46 of 40 to 50 percent could be achieved for the 125 mil ion knowledge workers in this category, which would lead to economic impact of $1.7 tril ion to $2.2 tril ion per year. A company cal ed SmartAction, for example, provides call automation solutions that use machine learning combined with advanced speech recognition to improve upon conventional interactive voice response (IVR) systems. SmartAction says that an auto club using its system has cut the time per cal for its 24/7 roadside assistance service by half, realizing cost savings of 60 to 80 percent over an outsourced cal center using human agents. These types of intel igent systems are able to automate many cal s while minimizing customer frustration with touch-tone or primitive voice-response systems (systems that offer prompts such as, “If this is correct, say yes”). This can lead to higher cal completion and lower abandoned cal rates.44 Currently available intel igent personal assistants such as Apple’s Siri and Google Now il ustrate some of the possible uses of these technologies in administrative support roles. The Google Now service already anticipates user needs, making recommendations or delivering information based on browser history, calendar entries, and current location. For example if traffic is bad, Google Now may suggest that the user leave early, having combined routing and traffic information with data about the time and location of the user’s next appointment. Social sector services Knowledge work automation could have important effects in education and health care, two large service sectors that are under pressure to improve productivity and quality. Knowledge work automation can augment teacher abilities and enhance or replace lectures with “adaptive” learning programs— dynamic instruction systems that alter the pace of teaching to match the student’s progress and suggest additional dril s based on student responses. Another area of potential impact is automated grading. A company cal ed Measurement Incorporated won a $100,000 prize from the Hewlett Foundation in 2012 for developing technology that enables a computer to grade student written responses, including essays, as wel as a skil ed human grader can.45 The economic impact of such tools in education would come from improving instructional quality and enabling teachers to provide more one-on-one attention and coaching. New self-teaching tools could also enable fundamental changes in scheduling: courses could be tied to subject mastery, rather than semesters or quarters, allowing students to progress at their own pace. In health care, oncologists at Memorial Sloan-Kettering Cancer Center in New York are using IBM’s Watson supercomputer to provide chronic care and cancer treatment diagnostics by accessing knowledge from 600,000 medical evidence reports, two mil ion pages of text from 42 medical journals, and 1.5 mil ion patient records and clinical trials in the field of oncology. It can then compare each patient’s individual symptoms, vital signs, family history, medications, genetic makeup, diet, and exercise routine to diagnose and recommend a treatment 44 Roadside assistance customers benefit from smart support during peak and after hours, SmartAction case study of Canadian Automobile Association Saskatchewan, June 2012. 45 Measurement Incorporated, “MI’s automated essay scoring system takes first prize in national competition,” press release, October 4, 2012.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 47 plan with the highest probability of success.46 This could be the first of many applications of knowledge work automation in medical diagnostics, given the costs of misdiagnoses. It is possible that productivity gains of 40 to 50 percent could be achieved by 2025 in the social sector categories we sized, which employ 55 mil ion people worldwide. We estimate that this could lead to economic impact of $1.1 tril ion to $1.4 tril ion per year. Technical professions The ability to use deep learning techniques to discover new relationships in huge amounts of data and to determine which relationships are the most important amounts to an enormous shortcut in many kinds of technical work—from software design to drug discovery. For example, by applying deep learning to drug development data, researchers can quickly narrow the field of possible formulations from thousands to dozens, drastical y speeding up the discovery process and saving thousands of hours of labor. A team of researchers in a contest sponsored by Merck recently proved that a deep-learning computer could examine an unfamiliar data set of chemical structures and develop its own rules to narrow down the thousands of unique molecules to those with the greatest potential to be effective.47 Software engineers are using machine learning to speed up software development through automated testing and algorithm performance optimization, as wel as project management tasks such as managing code libraries, tracking version control, and dividing tasks between developers.48 By 2025, there is potential for productivity gains of about 45 to 55 percent in this category, which employs about 35 mil ion knowledge workers worldwide. This could lead to economic impact of $1.0 tril ion to $1.2 tril ion per year. In our analysis, we only examined the impact of knowledge work automation on the cost of employment in technical fields, excluding the potential value of new scientific discoveries or improved research. Therefore, the ful impact of innovation in this category could be substantial y larger than our estimates. Management Machine learning excels at the complex analytics that managers use to monitor activities under their responsibility, understand the root causes of issues as they arise, and accurately forecast future trends on the horizon. For example, managers currently use machine-learning technology to monitor, control, and diagnose faults in manufacturing plants. By 2025, it is possible that productivity gains of 30 to 40 percent could be achieved for the 50 mil ion knowledge workers in this category, which would lead to economic impact of $0.8 tril ion to $1.1 tril ion per year. 46 Jonathan Cohn, “The robot wil see you now,” The Atlantic, February 20, 2013. 47 John Markoff, “Scientists see promise in deep-learning programs,” The New York Times, November 23, 2012. 48 Jitesh Dundas, “Machine learning helps software development,” Software Magazine, June 2012.
48 Professional services Fields such as law and financial services are already beginning to see the benefits of knowledge worker automation. Law firms, for example, are using computers that can scan thousands of legal briefs and precedents to assist in pretrial research—work that would once have taken hundreds or thousands of hours of paralegal labor. Symantec’s Clearwel system uses language analysis to identify general concepts in documents and present the results graphical y. In one case, this software was able to analyze and sort more than 570,000 documents in two days.49 Artificial intel igence (AI) has played a role in financial transactions for some time. AI algorithms are able to parse myriad news stories, financial announcements, and press releases, make decisions regarding their trading relevance, and then act in slivers of a second—faster and with greater information recal than any human trader.50 Banks can also use machine learning to detect fraud, finding charges or claims outside a person’s normal buying behavior. Even services like Future Advisor use AI to offer personalized financial advice inexpensively and at scale. Based on our estimates, it is possible that by 2025, productivity gains of 45 to 55 percent could be achieved for the 25 mil ion knowledge workers in this category, which would lead to economic impact of $0.6 tril ion to $0.8 tril ion per year. BARRIERS AND ENABLERS Realizing the ful potential impact of knowledge work automation wil involve overcoming some technological, regulatory, and organizational hurdles. Artificial intel igence, while showing remarkable advances, wil stil have to develop significantly before the scale of benefits that we estimate here can be realized. While we believe that very rapid advances in these areas could be possible over the coming decade, many of the most important future applications, such as diagnosis support for physicians, are stil in experimental stages. Cultural and organizational hurdles also exist. Risk-averse firms may delay adoption until the benefits of these technologies have been clearly proven. And, in some cases, there could be resistance. Many attorneys were initial y hesitant to use computerized research systems because they did not trust the machines to catch every document. Some business leaders might have concerns about legal liability regarding situations in which these technologies make mistakes (for example, with a patient diagnosis). Once the decision is made to adopt these technologies, business leaders wil need to prepare for how they wil be introduced, what tasks wil be either augmented or ful y automated, and how to alter roles and organizational processes to adjust for these changes. Figuring out how best to design roles and restructure organizations to ful y realize the value of this technology could take time. 49 John Markoff, “Armies of expensive lawyers replaced by cheaper software,” The New York Times, March 4, 2011. 50 For one interesting example of the use of AI in stock trading, see Christopher Mims, “AI that picks stocks better than the pros,” MIT Technology Review, June 10, 2010.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 49 Final y, in some cases there may be regulatory hurdles to overcome. To protect citizens, many knowledge work professions (including legal, medical, and auditing professions) are governed by strict regulatory requirements regarding who may perform certain types of work and the processes they use. Automated knowledge work applications in many highly regulated industries may need to undergo significant testing to verify their effectiveness before they wil be al owed to perform skil ed knowledge worker tasks; in many cases, humans may have to retain final review and approval over the work of these systems. IMPLICATIONS The automation of knowledge work has the potential to become pervasive, transforming the economics of many industries, but also posing chal enges and opportunities for technology providers, virtual y al business leaders, individuals, and policy makers. Technology providers (both software and hardware) wil play a critical role in this nascent field by designing powerful, easy-to-use knowledge work applications and supporting adoption within organizations. There could be many opportunities across a range of possible capabilities and approaches. Some technology providers might focus on high-end, advanced systems such as the Watson supercomputer, perhaps configured as enterprise solutions or programmed for specific verticals such as medicine. Others might focus on next-generation assistants similar to Apple’s Siri for both businesses and consumers. And many might focus on special-purpose tools for analytics, search functions, or a host of other potential applications. These knowledge work automation tools could be delivered in many ways, including via enterprise solutions, apps, or Web services. They could be delivered via the cloud (see Chapter 4) and on mobile Internet devices (see Chapter 1). They could also integrate with Internet of Things devices, both to analyze additional data and to directly control processes and environments (see Chapter 3). Many companies wil need support in change management, technical instal ation, process redesign, and employee training as they upgrade their technology platforms. Technology providers, IT consultants, and systems integrators are likely to find new opportunities to help businesses make these transitions successful y, perhaps using knowledge work automation technology themselves to better manage projects and conduct advanced analyses. The first task for business leaders is to understand how knowledge work is (and wil be) carried out in their organizations, including where the most time and money are spent, which functions contribute the most value, and which contribute the least, and where productivity is lowest. Answering such questions wil help set priorities regarding areas in which the adoption of tools to automate knowledge work might be both feasible and able to create consistently higher performance. Given the potential power of these tools, the biggest benefits may come from applying knowledge work automation to boost the productivity of employees in high-value-added functions, rather than focusing on simple tasks that might be turned over entirely to machines. To capture the benefits of knowledge work automation, companies and social service institutions wil need to manage fundamental organizational change. Many knowledge worker jobs could be redefined, and if so, workers wil need retraining, both to work with new technologies and to learn new tasks and skil s
50 as their jobs evolve. Some categories of knowledge jobs could become obsolete, as happened when word processing programs on desktop computers reduced the need for typists. In addition, much of the automation of knowledge work technology may require the intel igence of organizations to be codified, perhaps in many cases by the very workers who are adopting or even being replaced by this technology. This could create chal enges for employers looking to obtain robust employee support for adoption and wil require careful communication and change management. Knowledge work automation has the potential to provide enormous societal benefits, including helping to discover new medicines. These technologies could also directly address serious gaps in the supply of workers who have the skil s needed to drive 21st-century economies.51 However, the potential impact of this technology on employment could be a subject of intense debate. As we have seen in previous waves of manufacturing and transaction work automation, these changes often happen faster than social institutions can adjust. And while previous productivity gains have general y resulted in the emergence of new high- value-added jobs, it is not always the displaced workers who benefit most from these opportunities. Automation of knowledge work could drive the creation of many new types of jobs if businesses and governments can innovate effectively and adjust education and training to focus on new skil s. As with advanced robotics, these technologies could also create jobs for experts who can create and maintain the technology itself. However, increased productivity without innovation and retraining could ultimately exert downward pressure on wages and increase income disparities. The effects of these technologies on developing economies could be mixed. Some countries could lose opportunities to provide outsourced services if companies in advanced economies choose automation instead. But access to knowledge work automation technologies could also help level the playing field, enabling companies in developing countries to compete even more effectively in global markets. In addition to dealing with the employment and macro-economic effects of these technologies, policy makers and business leaders wil be confronted with legal and ethical considerations. How wil regulators and courts deal with harmful decisions made by computers (for example, if a computer were to give inappropriate medical treatment advice)? Who would be liable in such situations? Organizations might require that a human always make or approve final decisions, but what would happen when decisions and analyses become so complex as to exceed most people’s ability to ful y understand or audit them? We have already seen complex but poorly understood computer algorithms drive stock market turbulence. Similar risks could very wel arise in other applications. 51 See The world at work: Jobs, pay, and skil s for 3.5 bil ion people, McKinsey Global Institute, June 2012.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 51 Final y, as computers transform knowledge work in the coming decade, debates about the role of thinking machines in society wil undoubtedly intensify. Within this century, it could very wel be possible to create machines with processing powers that far exceed those of the human brain. What capabilities wil such machines have? How wil they be harnessed? Wil machines become “smarter” than humans? The answers to these questions wil no longer be left to science fiction writers, academics, and philosophers.
52 3. The Internet of Things Increasingly, the connected world includes physical objects. Machinery, shipments, infrastructure, and devices are being equipped with networked sensors and actuators that enable them to monitor their environment, report their status, receive instructions, and even take action based on the information they receive. Even people can be equipped with sensor-enabled devices—to track their health status, for example. This is what is meant by the term the Internet of Things, and it is growing rapidly. More than nine bil ion devices around the world are currently connected to the Internet, including computers and smartphones. That number is expected to increase dramatical y within the next decade, with estimates ranging from quintupling to 50 bil ion devices to reaching one tril ion.52 By bringing machines and assets such as shipping containers or hospital beds into the connected world, the Internet of Things enables new ways of monitoring and managing al the “moving parts” that make up a business. At any moment, management can see the status and flow of goods or materials through plants, distribution centers, and even onto store shelves. By monitoring machinery in real time, companies can better control the flow of goods through factories and avoid disruptions by taking immediate action or engaging in preventive maintenance when problems arise. Machines with embedded actuators in addition to sensors can be programmed to take action on their own. The widespread adoption of the Internet of Things wil take time, but that timeline is shrinking thanks to improvements in underlying technologies such as miniature sensors and wireless networks. The Internet of Things has the potential to create economic impact of $2.7 tril ion to $6.2 tril ion annual y by 2025. Some of the most promising uses are in health care, infrastructure, and public-sector services—helping society tackle some of its greatest chal enges. Remote monitoring, for example, has the potential to make a huge difference in the lives of people with chronic diseases while simultaneously attacking a significant source of rising health-care costs. The ability to monitor and control power grids and water systems could have major impacts on energy conservation, greenhouse gas emissions, and water loss. By using sensors to gather information to streamline operations, public-sector functions such as garbage col ection can become much more productive. Sensor data could also be used to improve policing. Realizing the ful potential of the Internet of Things wil not be easy. To capture the potential value of these applications, organizations wil need to have the systems and capabilities to make sense of the flood of data that remote sensors can provide. For example, with more widespread use of radio-frequency identification (RFID) tags, some companies could track hundreds of thousands, or perhaps even mil ions, of items in real time, requiring considerable analytical capabilities and talent. 52 Joseph Bradley, Joel Barbier, and Doug Handler, Embracing the Internet of everything to capture your share of $14.4 tril ion, Cisco Systems, February 12, 2013.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 53 Merging the physical and digital world also has implications for privacy, security, and even how companies are organized. As with any data connection, the connections that allow remote machines to take action without a human operator are subject to hacking by criminals or terrorists. The data col ected via health monitoring could be abused. Even the in-home control ers for some smart grid applications (for example, control ers that can selectively turn air-conditioning or appliances on and off to save energy or take advantage of lower rates) raise questions about privacy and autonomy. These issues wil need to be addressed before society and businesses wil be able to enjoy the ful benefits of the Internet of Things. DEFINITION The Internet of Things refers to the use of sensors, actuators, and data communications technology built into physical objects—from roadways to pacemakers—that enable those objects to be tracked, coordinated, or control ed across a data network or the Internet. There are three steps in Internet of Things applications: capturing data from the object (for example, simple location data or more complex information), aggregating that information across a data network, and acting on that information—taking immediate action or col ecting data over time to design process improvements. The Internet of Things can be used to create value in several ways. In addition to improving productivity in current operations, the Internet of Things can enable new types of products and services and new strategies: remote sensors, for example, make possible pay-as-you-go pricing models such as Zipcar. Internet of Things technology ranges from simple identification tags to complex sensors and actuators. RFID tags can be attached to almost any object. Sophisticated multisensor devices and actuators that communicate data regarding location, performance, environment, and condition are becoming more common. With newer technologies such as micro electromechanical systems (MEMS), it is becoming possible to place very sophisticated sensors in virtual y any object (and even in people). And, because they are manufactured using a semiconductor-like fabrication process, MEMS are rapidly fal ing in price. With increasingly sophisticated Internet of Things technologies becoming available, companies can not only track the flow of products or keep track of physical assets, but they can also manage the performance of individual machines and systems—an assembly line ful of robots and other machines, for example. Sensors can also be embedded in infrastructure; for example, magnetic sensors in roads can count vehicles passing by, enabling real-time adjustments in traffic signal timing. Equal y important as these sensors and actuators are the data communications links that transmit this data and the programming— including big-data analytics—that make sense of it al . Increasingly, Internet of Things applications include closed-loop setups in which actions can be triggered automatical y based on data picked up by sensors. For example, in process industries, sensor-based systems can automatical y react to incoming signals and adjust process flow accordingly. They can change a traffic light to green when a sensor in the pavement signals that cars are backed up at the intersection, or alert a doctor when the heart rate of a patient with a remote monitor spikes.
54 Basic uses of the Internet of Things are already wel under way. One of the biggest applications so far employs RFID to track the flow of raw materials, parts, and goods through production and distribution. These tags emit a radio signal that can be used to pinpoint their location. So, for example, as a tagged product moves through a factory, computers can track where it is at any given moment. Using that information, the company can spot bottlenecks, managing the timing of the release of more parts into the system, or schedule trucks to pick up finished goods. RFID tags on containers and boxes are used to track products as they make their way through warehouses and transportation hubs to store shelves and (in cases where tags are used on packaging) even al the way to the consumer. Tracking these flows gives companies the opportunity to tighten supply chains and avoid stock-outs or building too much inventory. RFID tags are also used in E-ZPass tol -col ection systems, speeding traffic flow on tol roads and bridges. In another example, FedEx now offers a program that al ows customers to track the progress of packages almost continuously by placing a smal device (about the size of a mobile phone) into packages. These devices contain both a global positioning system and sensors to monitor temperature, humidity, barometric pressure, and light exposure, which are critical to cargo such as biological samples and sensitive electronic equipment. These devices are programmed to relay location and atmospheric condition information periodical y so customers can know the exact whereabouts and condition of their packages and learn immediately if they veer off course or have been exposed to risky conditions. This type of continuous data availability obviously has implications for companies that operate long and complex supply chains. POTENTIAL FOR ACCELERATION The Internet of Things is stil in early stages of adoption, but it already has a wide variety of uses, and the portfolio of applications is expanding daily. As Internet of Things technology proliferates, it has the potential to address many major needs, including improved resource productivity and infrastructure management. Smart grids for electricity, water, and transportation networks are examples. Electric and water utilities have been among the early adopter industries. Sensors are essential to smart grid systems, which give utility operators a way to gauge usage and network performance in real time. This means that rather than waiting to receive cal s from customers whose lights have gone out, the electric company can spot a failure as it happens and, under some circumstances, even restore power by rerouting service around the failed transmission or generating equipment. Minnesota Power, a US utility company, has instal ed a smart grid system and upgraded feeder lines that al ow the company to offer “100 percent uptime” to commercial customers.53 Internet-connected sensors are also being used to take seismic readings under the earth’s crust and monitor the flow of water through supply pipes. In the energy industry, sensors are used to map unexplored fossil fuel fields to pinpoint deposit locations. 53 SGIG accelerates grid modernization in Minnesota, Smartgrid.gov, US Department of Energy.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 55 Internet of Things technology can also have a direct impact on human lives and health. The so-cal ed Quantified Self concept—which involves using sensors to track exercise performance or monitor health—is an increasingly popular trend powered by Internet of Things technologies. For example, several companies are now sel ing wearable sensors that al ow consumers to track the number of miles they run, their heart rate, and other data generated during exercise, which can then be used to manage health. Doctors now perform “capsule endoscopy” using a pil -shaped micro-camera with wireless data communication capabilities that travels through a patient’s digestive system and transmits images to a computer. Several technological advances are improving the effectiveness of Internet of Things applications while also reducing costs. The price of RFID tags and sensors is falling, and new developments such as MEMS are enabling new uses. Sales of sensors have grown by 70 percent annual y since 2010, and advances in technology are making more capable sensors more affordable. More sensors of various types are being integrated into more physical devices, and improved power management is al owing devices to run unattended for longer periods of time. Miniaturization and high-volume manufacturing techniques make it possible to instal sensors in even the smal est devices; for example, a smartphone may have a single chip that includes a positioning sensor, a thermometer, and a motion detector. Final y, the spread of high-speed wireless data networks is extending the coverage area of the mobile Internet, helping pave the way to greater Internet of Things uses. POTENTIAL ECONOMIC IMPACT BY 2025 We estimate the potential economic impact of the Internet of Things to be $2.7 tril ion to $6.2 tril ion per year by 2025 through use in a half-dozen major applications that we have sized (Exhibit 5). The largest impacts among sized applications would be in health care and manufacturing. Across the health-care applications we analyzed, Internet of Things technology could have an economic impact of $1.1 trillion to $2.5 trillion per year by 2025. The greatest benefits in health care could come from improved efficiency in treating patients with chronic conditions. Using sensors that read the vital signs of patients at home, nurses and doctors can be alerted to emerging problems, such as a dangerous drop in the glucose levels of a diabetic patient. Advising patients about how to address problems at home or treating them in outpatient settings lowers the frequency of costly emergency room visits and unnecessary hospitalizations. Treatment costs for chronic diseases constitute approximately 60 percent of total health-care spending, and the annual cost of these diseases in 2025 could be as high as $15.5 tril ion global y.54 We estimate that remote monitoring could reduce this cost by 10 to 20 percent where applied, although realized value might be reduced by factors such as adoption rates and patient acceptance (or resistance).55 54 McKinsey estimate based on current data from Canada, France, the United States, and the United Kingdom. 55 Based on a case study by the US Veterans Health Administration regarding chronic heart failure, diabetes, and chronic obstructive pulmonary disease, including more than 70,000 patients. See Andrew Broderick and David Lindeman, “Scaling telehealth programs,” Case Studies in Telehealth Adoption, January 2013.
Exhibit 5 Sized applications of the Internet of Things could have direct economic impact of $2.7 trillion to $6.2 trillion per year in 2025 Potential economic impact of sized Sized applications in 2025 Estimated scope Estimated potential Potential productivity or applications $ tril ion, annual y in 2025 reach in 2025 value gains in 2025 $15.5 tril ion cost of 70–80% mobile 10–20% cost reduction in treating chronic penetration in patients chronic disease diseases who account for bulk of treatment through $400 bil ion cost of health-care spending remote health monitoring counterfeit drugs, Counterfeit drug tracking 80–100% reduction in 40% addressable – Developed world: drug counterfeiting 1.1– with sensors 50–80% 0.5–1.0 hour time saved Health care 2.5 50 mil ion nurses for – Developing world: per day by nurses inpatient monitoring 20–50% – Developed world: Inpatient monitoring $30 per hour – Developed world: Manufac- 0.9– – Developing: 75–100% turing 2.3 $15 per hour – Developing: 0–50%
$47 tril ion in global 80–100% of al 2.5–5.0% saving in manufacturing manufacturing operating costs,
0.2– Electricity operating costs including maintenance
0.5 and input efficiencies 27,000–31,000 TWh 25–50% of consumers 2–4% reduction in Urban global electricity could adopt energy demand peaks in the grid 0.1– infra- consumption management Reduction of total load 0.3 structure $200 bil ion spending 25–50% of grid on grid on transmission lines monitored through Operating/maintenance 300 bil ion consumer sensors savings; shorter outage minutes outage 50–100% of consumer time through automated 0.1– Security meters automated meters 0.2 200–300 hours 40–70% of working urban 10–20% reduction in commuting time per population living in cities average travel time urban worker with smart infrastructure through traffic and Resource 0.1– per year 50–70% of large urban congestion control extraction 0.2 $200 bil ion spent on regions adopting smart 10–20% reduction in urban water water infrastructure and water consumption and $375 bil ion cost of waste handling leaks with smart meters waste handling and demand control Agriculture ~0.1 10–20% reduction in cost of waste handling $6 tril ion cost of Adoption of advanced 4–5% crime reduction 0.02– crime surveil ance by countries through improved Retail 0.10 accounting for 50–70% of surveil ance global GDP $3.7 tril ion in global 80–100% of al resource 5–10% saving in mining operating extraction operating costs from Vehicles ~0.05 costs productivity gains $630 bil ion in 10–30% of al insured 25% reduction in cost of Other automotive insurance cars equipped with vehicle damage from potential premiums1 sensors col ision avoidance and applications
increased security1 (not sized) $200 bil ion lost due 30–80% of retail adopting 1.5–2.0% increased to stockouts smart logistics sales Sum of sized potential 2.7– $1.2–1.3 tril ion in 20–40% adoption of 10–20% increase in economic 6.2 agricultural advanced irrigation yields from precision impacts production (wheat, systems and precision application of fertilizer maize, rice, farming and irrigation soybeans, barley) 1 Automotive premiums used as proxy for cost of col isions. NOTE: Estimates of potential economic impact are for some applications only and are not comprehensive estimates of total potential impact. Estimates include consumer surplus and cannot be related to potential company revenue, market size, or GDP impact. We do not size possible surplus shifts among companies and industries, or between companies and consumers. These estimates are not risk- or probability-adjusted. Numbers may not sum due to rounding. SOURCE: McKinsey Global Institute analysis 5
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 57 Additional value from use of Internet of Things systems in health care would include in-hospital health monitoring. Based on cases where physicians and nurses have had access to real-time patient data, potential gains of 30 minutes to one hour of time per day per nurse could be possible. Counterfeit drugs are another health-care problem with a possible Internet of Things solution. Currently, more than $75 bil ion worth of counterfeit drugs are sold per year, and that amount is growing by around 20 percent annual y.56 Using sensors on bottles and packages could reduce the sale of counterfeit drugs by enabling consumers to ensure that their medications are legitimate. We estimate that this technique could apply to 30 to 50 percent of drugs sold and could be successful 80 to 100 percent of the time. In manufacturing, Internet of Things technology can improve operational efficiency in a variety of ways. Sensors can be used to track machinery and provide real- time updates on equipment statuses, decreasing downtime. Sensors can also be placed on trucks and pal ets to improve supply chain tracking and management. They can be used to monitor the flow of inventory around factory floors or between different workstations, reducing work-in-progress inventory levels, decreasing wait times, and creating transparency to better optimize flows. In precision manufacturing, sensors and actuators can even be used to change the position of objects as they move down assembly lines, ensuring that they arrive at machine tools in an optimum position, avoiding the smal deviations in the position of work in process that can jam or even damage machine tools. We estimate that productivity gains equivalent to 2.5 to 5 percent are possible from Internet of Things applications that we sized in discrete and process manufacturing industries. The total operating cost of global manufacturing is currently about $25 tril ion per year, and could reach more than $47 tril ion by 2025. Given the low cost of sensors and the large demand for process optimization in manufacturing, very high adoption rates are possible; in fact, perhaps 80 to 100 percent of al manufacturing could be using Internet of Things applications by 2025. This would lead to potential economic impact of $900 bil ion to $2.3 tril ion per year by 2025. Smart electrical grid systems are an important Internet of Things application, with a potential annual value that we estimate could be $200 bil ion to $500 bil ion by 2025. The bulk of this impact would come from demand-management applications that could reduce costly peak usage, which often requires utilities to buy electricity at the highest rates or invest in extra peak capacity (see Chapter 8, “Energy storage”). Many commercial customers already avoid scheduling energy- intensive processes and production during periods of peak energy demand, when energy costs are at their highest, and some have formal agreements with utilities to reduce usage whenever demand reaches a certain level. Operators of data centers, which constitute one of the fastest-growing consumers of electricity, are starting to adopt power-management techniques based on real-time grid information. With smart grids, consumers can let the utility company automatical y power down high-use systems and appliances during periods of peak demand or they can make their own choices based on real-time rate information that the utility supplies. Demand management could reduce peak demand by 2 to 4 percent and cut overal demand by 1 to 2 percent. This would al ow utilities 56 Center for Medicine in the Public Interest.
58 to avoid building potential y bil ions of dol ars’ worth of additional capacity and infrastructure. Smart grids also help cut utility operating costs by providing real-time information about the state of the grid. Potential benefits of this include reducing total outage times and decreasing electricity waste by better regulating voltage and balancing load between lines. Grid sensors can monitor and diagnose network problems to prevent outages and reduce maintenance costs. At the user end, smart meters equipped with two-way communication capabilities could reduce outage minutes and enable faster outage detection. Smart meters also enable automatic meter reading, eliminating the need for personnel to gather that information. The Internet of Things is an important enabler of better management of urban infrastructure, systems, and services, including traffic, waste and water systems, and public safety. Sensors that monitor traffic patterns can generate the data to optimize flow by adjusting traffic light timing, imposing congestion charges, and changing bus routes, for example. Sensors can automatical y trigger an alert to divert traffic around accidents to minimize costly delays. London, Singapore, and Houston have al realized significant reductions in commuting times using this technology. Based on these examples, cities could cut motor vehicle commuting time by 10 to 20 percent on average, saving hundreds of mil ions of hours a year. Cities can also use Internet of Things technology to streamline garbage col ection and improve water management. In the United States, the cities of Cleveland and Cincinnati in Ohio have both supplied households with garbage and recycling bins equipped with RFID tags, which al ow city crews to see whether residents are putting out garbage and recycling on the designated days. As a result of these data, Cleveland was able to eliminate 10 pickup routes and cut operating cost by 13 percent through improved labor productivity. Both cities also instituted “pay as you throw” programs, which require residents to pay extra for putting out more garbage than fits in city-issued bins. In Cincinnati, residential waste volume fell 17 percent and recycling volume grew by 49 percent through the use of these programs. Using conservative assumptions, such measures could reduce waste handling costs by 10 to 20 percent by 2025. The cities of Doha, São Paulo, and Beijing al use sensors on pipes, pumps, and other water infrastructure to monitor conditions and manage water loss, identifying and repairing leaks or changing pressure as necessary. On average, these cities have reduced leaks by 40 to 50 percent. Smart meters at the user end al ow real-time monitoring of demand and leak detection by residents and property managers, reducing costs. Dubuque and Indianapolis in the United States, as wel as Malta, New Delhi, and Barrie (Ontario), have seen, on average, a 5 to 10 percent reduction in water usage via the use of smart water meters. The total potential economic impact from traffic applications, smart waste handling, and smart water systems in urban areas could total $100 bil ion to $300 bil ion per year by 2025. This assumes that 80 to 100 percent of cities in advanced economies and 25 to 50 percent of cities in the developing world could have access to this technology by that time. The Internet of Things can also improve law enforcement efforts. It wil soon be possible to place inexpensive sensors on light poles, sidewalks, and other objects on public property to capture sound and images that can be analyzed
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 59 in real time—for example, to determine the source of a gunshot by analyzing the sound from multiple sensors. This could potential y take policing to a new level of effectiveness, creating an opportunity to reduce both the human and economic costs of crime. The economic cost of crime is estimated to be 5 to 10 percent of GDP around the world. If 4 to 5 percent of this could be eliminated, the potential economic impact could be $100 bil ion to $200 bil ion per year in 2025. In the oil, metal, and mineral extraction industries, Internet of Things technology could help find and map mineral deposits and increase recoverability. Operating costs reductions of 5 to 10 percent have been realized through the use of sensors and big data in basic material extraction. The total operating cost for the oil, metal, and mineral extraction industries in 2025 is estimated to be $1.4 tril ion. The adoption of Internet of Things technologies could be very high in this industry, perhaps 80 to 100 percent. At that level of adoption, potential economic impact of $100 bil ion to $200 bil ion per year might be possible by 2025. In agriculture, the Internet of Things has the potential to create significant value. For example, leaf sensors can measure stress in plants based on moisture levels. Soil sensors can gather information about how water moves through a field and track changes in soil moisture, carbon, nitrogen, and soil temperature. Such data would help farmers optimize irrigation schedules, avoiding crop damage. Soil and plant data can be used to guide “drip-fertigation,” which applies liquid fertilizer through drip irrigation systems to ensure crops receive the correct amount of nutrients and water at al times. For example, in the United States, drip fertigation was used by Stamp Farms in Decatur, Michigan, to increase yields by 10 to 40 percent in corn. We estimate that using sensor data for “precision farming” could raise yields 10 to 20 percent global y. Assuming that 25 to 50 percent of farms adopt this approach, we estimate that in these applications, the Internet of Things could have the potential to create $100 bil ion per year in economic impact in 2025. The Internet of Things could help address the out-of-stock chal enge in retail sales. It is estimated that retailers lose the equivalent of 4 percent of sales every year due to items desired by the consumer that are not in stock. By 2025, this could represent $200 bil ion a year in lost value. We estimate that 35 to 50 percent of this value can be recaptured by using sensors and tags to tighten supply chains and predict where stock-outs are likely to occur. This could drive potential economic impact of $20 bil ion to $100 bil ion per year by 2025. Adding sensors to automobiles to prevent crashes could create economic value of as much as $50 bil ion per year by 2025. This estimate is based on the reduced property damage that would occur if automatic braking systems were widely used and prevented a large portion of low-speed col isions (we do not consider high- speed col isions, which often involve injury or death, in this analysis). We estimate that 25 percent of the damage caused by low-speed accidents could be avoided using Internet of Things technology, potential y resulting in $50 bil ion global y in reduced property damages. BARRIERS AND ENABLERS The Internet of Things offers great promise, but al the pieces are not yet in place to guarantee that rising interest wil turn into widespread investment and adoption. There are technical, financial, and regulatory issues that must be resolved. For
60 example, early adopters wil need to prove that sensor-driven business models create superior value. On the technology side, the cost of sensors and actuators must fal to levels that wil spark widespread use. Also, technology providers need to agree on standards that wil enable interoperability between sensors, computers, and actuators. Until such standards exist, investing in Internet of Things applications wil require extra effort to build and maintain integrated systems. It wil also carry the additional risk of betting on the wrong technology, which could slow adoption. Progress is also needed in creating software that can aggregate and analyze data and convey complex findings in ways that make them useful for human decision makers or for use by automated systems (for example, calculating medication dosages based on real-time patient data). The Internet of Things also faces hurdles due to privacy and security concerns, which wil require action by both businesses and policy makers. As Internet of Things applications become more sophisticated and more operations fal under the supervision of sensor-based systems, data security and network reliability will be important concerns. As sensors are introduced into the lives of consumers via traffic control systems, health-care applications, smart grids, and retail space uses, concerns are likely to grow over how the data that are col ected wil be used. Wil the information from medical monitors be used to deny individuals health insurance coverage? Could hackers steal sensor data regarding how your car moves in order to track your personal movements? Both businesses and regulators wil have to address questions such as this to foster widespread adoption of these technologies. For both consumers and businesses, sensor-based systems also create liability issues that policy makers need to address. For example, it is not ful y clear who wil be legal y responsible for injuries or damages that are caused by errors in closed-loop systems in which an algorithm dictates the actions of a machine. IMPLICATIONS The Internet of Things is such a sweeping concept that it is a chal enge to even imagine al the possible ways in which it wil affect businesses, economies, and society. For the first time, computers are now able to receive data from almost any kind of physical object, enabling us to monitor the wel -being and performance of machines, objects, land, and even people. Using the data from these sources, computer systems wil be able to control machines, manage traffic, or tel a diabetic it is time to eat. Businesses wil be chal enged to make the most effective use of this technology given the level of innovation and technical expertise that wil be required. This is new territory for almost everyone, even those with a high degree of technical expertise. Policy makers wil likely have a long list of issues to resolve to al ow the benefits of Internet of Things applications while protecting the rights and privacy of citizens. For technology suppliers and the companies that adopt that technology, the Internet of Things promises rewards that wil not always be easy to obtain. Hardware manufacturers who supply sensors, actuators, and communications devices wil be pressed to continue to refine their products and reduce costs. For example, despite many years on the market, RFID tags are stil too expensive for many businesses to use as extensively as was predicted a decade ago. Moreover, because of the complexity of systems that could require hundreds of thousands
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 61 of devices, sensors, and other hardware wil need to be reliable, maintenance- free, and interoperable. New partnerships wil be needed between companies with capabilities in sensors and manufacturers of the machines, products, and objects into which they wil go. Some of the best-positioned companies may be suppliers of big data and analytical software that can help extract meaning from the enormous flows of data that the Internet of Things wil produce. Companies that hope to reap the benefits of operational improvements and use the Internet of Things to deliver new kinds of customer service and higher-quality products wil face an array of technological and organizational chal enges. Over the past two decades, the need to understand and use IT tools has spread across organizations. The Internet has forced sales and marketing departments to become masters of websites and Web analytics, for example. The Internet of Things takes this trend to a new extreme in which every department within an organization, from production to logistics to customer service and sales, could potential y receive real-time data about how the company’s products are being built, distributed, sold, and used. Few organizations are ready to deal with this sheer amount of data and have personnel who are able to do so. Gaining access to the talent to manage Internet of Things applications and educating executives and managers across functions wil need to be a top priority. For policy makers, the Internet of Things also brings great opportunities and chal enges. As operators of infrastructure and public services (often including health care), governments could be major adopters of Internet of Things applications. These technologies can reduce costs and improve quality of service, sometimes in lifesaving ways. City residents could see traffic flowing more smoothly, garbage picked up more efficiently, crime reduced, and water systems operating more efficient. The potential is enormous—but as in business, it wil not be realized without substantial investments in capabilities. In terms of public policy, government leaders wil need to establish clear understandings of the privacy risks that accompany the Internet of Things. The ability to put sensors virtual y anywhere—to observe the traffic on a residential street or to monitor a home’s electricity use—wil undoubtedly raise serious concerns about how al that information wil be used. Realizing the benefits of the Internet of Things in policing, for example, may require an unprecedented level of surveil ance that the public may reject. Policy makers faced with these issues wil need to think comprehensively and global y. One-off regulations and rules that are in conflict from one jurisdiction to another wil not suffice. Policy makers wil need to build consensus regarding what protections to put in place and work across borders and levels of government to make sure these protections can and wil be universal y enforced. Unfortunately, computer systems and networks could be the targets of criminals, terrorists, or even just hackers trying to prove a point. With sensors and networks control ing critical systems such as the electric grid, the consequences of such attacks could be staggering. It wil take a great deal of thought and planning, as wel as col aboration with the private sector, to both create proper safeguards and keep them up to date as technological advances continue.
62 4. Cloud technology Cloud technology has become a huge buzzword in recent years, and with good reason. The cloud already creates tremendous value for consumers and businesses by making the digital world simpler, faster, more powerful, and more efficient. In addition to delivering valuable Internet-based services and applications, the cloud can provide a more productive and flexible way for companies to manage their IT. Cloud technology has the potential to disrupt entire business models, giving rise to new approaches that are asset-light, highly mobile, and flexible. Cloud technology al ows the delivery of potential y al computer applications and services through networks or the Internet. With cloud resources, the bulk of computational work can be done remotely and delivered online, potential y reducing the need for storage and processing power on local computers and devices.57 The most commonly used Internet services are already delivered through the cloud, including online searching, social networks, and streaming media. The cloud also enables pay-as-you-go models for consuming IT, as exemplified by the phrases “software as a service” and “infrastructure as a service.” By 2025 most IT and Web applications and services could be cloud- delivered or -enabled, and most businesses could be using cloud facilities and services for their computing resources. The cloud enables some of the most highly impactful technologies we analyze in this report: mobile Internet, automation of knowledge work, and the Internet of Things. Since apps often rely on cloud resources, the cloud is expected to be a major driver of smartphone use. The total economic impact of cloud technology could be $1.7 tril ion to $6.2 tril ion annual y in 2025. Most of this impact ($1.2 tril ion to $5.5 tril ion) could be in the form of additional surplus generated from cloud delivery of services and applications to Internet users, while the rest could result from the use of cloud technology to improve enterprise IT productivity. As the cloud setup becomes a dominant computing paradigm, it could have wide-ranging implications for businesses, consumers, and policy makers. Consumers wil likely continue to benefit as new cloud-enabled apps and services emerge and reduce the need to instal and maintain local applications. Providers of “public” cloud services (services offered to multiple businesses) could see new competition from both large technology companies and their current enterprise customers, who could decide to develop their own cloud capabilities. Enterprises that take advantage of public or private cloud models could potential y see productivity gains and enjoy increased flexibility. Small enterprises and entrepreneurs could be able to use the agility provided by cloud technology to level the playing field with larger rivals. Final y, as cloud technology enables Internet-based delivery of more and more applications and services, 57 This concept is also sometimes referred to as having a “thin client.”
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 63 policy makers wil be under pressure to update laws relating to data ownership and privacy as they relate to the cloud. DEFINITION The concept of the cloud originated as a symbol for networks on diagrams of IT systems. Back then—in the days before personal computers, which first made processing and storage on the desktop the norm—users had “dumb” terminals that relied on a minicomputer or mainframe somewhere on the network (the cloud) for al resources. The modern cloud brings computer architecture full circle, enabling network access (from a computer or mobile Internet device) to a shared pool of computing resources such as servers, storage, and applications that can be used as needed.58 Behind the scenes, this requires a complex system of servers and storage systems that can al ocate computing resources to serve multiple customers simultaneously and keep track of what each user needs. This is the technology that makes it possible for a consumer to begin streaming a movie on a PC, pause it, then resume it from a tablet. When thousands of users suddenly demand the same content, streaming services seamlessly tap more processing power, then release the excess capacity when demand fal s below the peak. In enterprise IT, cloud technology provides on-demand self-service, anytime and anywhere availability, pooling of computing resources for multiple users or organizations, and usage-based pricing. One of the chief advantages of the cloud model is elasticity—users can expand or shrink capacity as needed. Cloud technology can be implemented as a third-party service or by companies that pool their computing resources on their own private clouds. By centralizing computers, storage, and applications on the cloud, companies raise IT productivity by increasing utilization (which is currently limited by the fact that many computers are used at peak capacity for only 30 to 40 days a year) and reducing the number of employees needed to maintain systems and develop software. With public clouds, companies can move to an “asset-light” model by turning a large capital investment (IT infrastructure) into an operating cost. Cloud setups are also more reliable (since they are capable of shifting processing from one machine to another if one becomes overloaded or fails), eliminating productivity-draining outages. POTENTIAL FOR ACCELERATION The biggest driver of incremental cloud technology demand in the coming decade could be the rapid proliferation of services and applications for Internet “clients”— the computers and mobile devices that are used to access online services and resources. The world population of Internet users is estimated at about 2.5 bil ion today, and could swel to more than 5 bil ion by 2025 thanks to the rapid proliferation of smartphones. Not only wil there be more Internet users in the near future, but these users wil also be relying more on off-device processing, storage, and applications. Consumers are using mobile Internet devices for increasingly 58 The ful definition according to the US National Institute of Standards and Technology: “Cloud computing is a model for enabling ubiquitous, convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction.” See Peter Mel and Timothy Grace, The NIST definition of cloud computing, National Institute of Standards and Technology, NIST Special Publication 800-145, September 2011.
64 demanding applications, including HD video streaming. Al of that computational work wil likely be carried out on cloud systems. Demand for enterprise cloud services wil continue to grow as wel . IT departments face ever-growing demands to do more and improve productivity at the same time. Cloud computing not only cuts costs, but also helps companies implement new applications and add services and computational capacity more quickly than they can using in-house staff. Another source of cloud usage growth could be smal and medium-sized enterprises (SMEs), which may have even more to gain from cloud services than do large corporations. Smal companies often find it difficult to build and manage extensive IT infrastructure and plan for future needs. Like larger enterprises, they also often struggle with a poor rate of return on IT systems due to rapid obsolescence of technology. Cloud computing lets SMEs avoid tying up capital in IT and frees them from IT infrastructure management and demand planning, giving them the ability to compete more effectively with big companies. The utility of the cloud extends to software and applications (so-cal ed software as services). For example, Microsoft Office 365 and Google Apps offer suites of applications available over the Internet (instead of via traditional software packages that must be purchased and instal ed). Meanwhile, the cost of implementing cloud setups has fal en, while performance has improved. For example, renting a server in the cloud is now about one-third as expensive as buying and maintaining similar equipment. According to the Cisco Global Cloud Index, global cloud traffic could increase by a factor of six in the next five years; by 2019, more than two-thirds of the global traffic through data centers could be cloud-based—double what it is today.59 POTENTIAL IMPACT BY 2025 While enterprise IT use wil continue to grow, the largest source of economic impact through 2025 wil likely come from enabling the delivery of services and applications to Internet users. We estimate the total potential economic impact for cloud technology across sized applications could be $1.7 tril ion to $6.2 tril ion in 2025 (Exhibit 6). Of this total, $1.2 tril ion to $5.5 tril ion could be in the form of surplus from use of cloud-enabled Internet services, while $500 bil ion to $700 bil ion could come through productivity improvements for enterprise IT. In estimating the potential incremental consumer surplus from cloud computing, we assume the Internet wil continue to grow at projected rates through 2025 and that users wil continue to use email and social networks, entertainment services (music, video, and games), and Web services such as search and mapping. We used research on Internet surplus by McKinsey and IAB Europe to estimate the current surplus.60 We have used a surplus per user growth rate based on estimates of the growth rate of time spent by a typical user on the Internet. At the low end of our range, we assume time spent could grow at current rates and nearly double by 2025; at the high end, we assume nearly al media consumption 59 Cisco Global Cloud Index: Forecast and Methodology, 2011–2016, Cisco Systems, October 2012. 60 Consumers driving the digital uptake: The economic value of online advertising-based services for consumers, McKinsey & Co. and IAB Europe, September 2010; Internet matters: The Net’s sweeping impact on growth, jobs, and prosperity, McKinsey Global Institute, May 2011.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 65 could eventually be provided through cloud-enabled Internet services.61 We apply this estimate of surplus on an estimated 2 bil ion to 3 bil ion additional Internet users, as wel as 2.5 bil ion existing users who wil gain incremental utility. Developing economies could be home to more than 3.5 bil ion Internet users by 2025, many of whom may have only mobile devices (see Chapter 1: Mobile Internet).62 We assume, however, that the surplus available from the use of mobile Internet wil be lower than that from other forms of Internet access.
Exhibit 6 Sized applications of cloud technology could have economic impact of $1.7 trillion to $6.2 trillion per year in 2025 Potential economic impact of sized applications in 2025 Estimated scope Estimated potential Potential productivity Sized applications $ tril ion, annual y in 2025 reach in 2025 or value gains in 2025 2–3 bil ion more Nearly al Internet $25–85 surplus per Internet users, applications use cloud user per month Surplus from 1.2– most in developing as a core enabler cloud-based 5.5 economies Internet $1.26 tril ion or 20–30% productivity 40% of global IT Varying levels of gains
Infrastructure cloud adoption 0.3– spending in base Reduced
and operating across enterprises 0.4 scenario2 infrastructure and
expenses All enterprises facilities footprint Enterprise could have Higher task productivity1 potential to use Application standardization and cloud development 0.2– automation Most enterprises and packaged 0.3 $1.68 tril ion or may use a hybrid 10–15% productivity software 60% of global IT cloud gains spending in base The share of Standardization of Other scenario2 public cloud application potential usage may environment and applications increase as packages (not sized) cybersecurity Faster improves experimentation Sum of sized and testing potential 1.7– economic 6.2 impacts 1 We have not sized the impact of increased flexibility and convenience to enterprises. 2 Estimates for enterprise cloud based on a global IT budget that does not include telecommunications. NOTE: Estimates of potential economic impact are for some applications only and are not comprehensive estimates of total potential impact. Estimates include consumer surplus and cannot be related to potential company revenue, market size, or GDP impact. We do not size possible surplus shifts among companies and industries, or between companies and consumers. These estimates are not risk- or probability-adjusted. Numbers may not sum due to rounding. SOURCE: McKinsey Global Institute analysis To size the potential for productivity improvements due to cloud technology, we built a hypothetical “base” scenario of global IT spending and then estimated potential productivity improvements using this base across two main categories: infrastructure (both capital and operating expenses) and software development and packaged software costs. To construct this base scenario, we assumed a growth rate for each of these categories informed by historical rates and other factors, such as standardization of software, without assuming discontinuities in cloud adoption. 61 Lisa E. Phil ips, “Trends in Consumers’ Time Spent with Media,” eMarketer.com, December 28, 2010 62 Several economists have tried to size the surplus generated by the Internet. See Hal Varian, “The value of the Internet now and in the future,” The Economist, March 10, 2013; Shane Greenstein, “Measuring consumer surplus online,” The Economist, March 2013; Consumers driving the digital uptake: The economic value of online advertising-based services for consumers, McKinsey & Co. and IAB Europe, September 2010. 6
66 We then applied an estimate for productivity improvement in this base scenario to obtain a range for potential impact from cloud adoption by enterprise IT. For example, we have applied a potential productivity improvement of 20 to 30 percent in infrastructure support activities based on the efficiencies that arise when data centers are consolidated. For software development and packages, our estimate is 10 to 15 percent productivity improvements created by time savings and quality improvements. Together, this leads us to productivity improvements of $500 bil ion to $700 bil ion by 2025. However, given the IT needs across al sectors in the global economy, as wel as the impact of greater flexibility and agility for which we have not estimated economic value, the actual impact could be even greater. BARRIERS AND ENABLERS Network capacity is a critical enabler for cloud computing, especial y wireless networks for consumers using the mobile Internet. Cloud technology is deployed through massive data centers that require high-capacity bandwidth. Some mobile networks are already straining to keep up with ever-increasing demand. The lag between rising demand and network capacity expansion could grow even wider as consumers try to stream more HD content and businesses begin to monitor mil ions of devices over wireless Internet services. Some companies are already starting to trial the next generation of broadband networks to anticipate and meet this demand—for example, Google Fiber in Kansas City, which is nearly 100 times as fast as current commercial y available high-speed broadband.63 Reservations by some consumers and enterprises regarding trusting the cloud represent a potential y significant hurdle. The cloud requires a level of trust that some managers and consumers are reluctant to grant. Many consumers still prefer to store their data on PC hard disks instead of trusting the cloud as a permanent and secure repository for their photos, personal records, and other irreplaceable material. Enterprises, too, continue to have concerns about placing sensitive data on a third-party cloud, especial y as questions of ownership and liability for data residing in a particular online location have yet to be settled by policy makers. Despite improvements in cloud technology, high-profile failures continue, affecting public perception of cloud reliability. For example, Amazon Web Services suffered an outage on Christmas Eve in 2012, taking down popular services such as Netflix for almost a day.64 Structural issues and cultural resistance in IT departments are also barriers. Cloud deployment represents a significant shift in IT management practices, from in-house work to lower-cost, outsourced solutions, raising concerns about loss of control. Moving to cloud technology also requires new IT budget discipline and skil sets. Another factor is the complexity of migrating enterprise IT systems with multiple platforms, network protocols, and programming environments to the cloud. 63 Greg Kumparak, “Google announces Provo, Utah as the third Google Fiber city and acquires the local fiber provider,” TechCrunch.com, April 17, 2013. 64 Stephen Musil, “Amazon apologizes for Netflix’s Christmas Eve streaming outage,” CNet.com, December 31, 2012.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 67 IMPLICATIONS As a core enabler of the Internet, cloud computing could have a huge impact on consumers’ lives. The surplus generated by the Internet could be valued in tril ions of dol ars by 2025. The imperative for lighter and faster mobile devices as the primary means of using online services, combined with consumption of rich media such as movies and music, could make cloud computing a popular (and nearly default) technology for delivering applications and software as services. The rise of the Internet of Things, exponential growth in data, and increasingly automated knowledge work could al rely on cloud technology to provide the computational resources needed to harness their benefits. Though we expect growing adoption of cloud technology as a computing paradigm for enterprises, stakeholders across the provider value chain could be differently affected. True, demand for cloud solutions could increase; however, providers of integrated cloud services may face greater competition as large technology players join early leaders in offering services. Many of these companies build their own data centers and cloud applications from the bottom up, buying components directly from manufacturers and then customizing completely, and could view public cloud services as a natural extension of their experience in building data centers and running private clouds. At the same time, manufacturers of components (such as motherboards, chipsets, and router switches) could recognize the potential for sel ing directly to enterprise customers keen to customize to their requirements. These manufacturers could begin to invest in building customer-facing sales organizations, potential y competing against server manufacturers for business from cloud providers. More software and media companies can be expected to transition from sel ing products to offering cloud-based services, which has implications for their business models. A sel er of packaged software or solutions may achieve lower margins by switching to a software-as-a-service model, but the cloud-based model could attract new customers by making products more affordable and accessible. For example, renting a movie on a subscription-based cloud service costs much less than renting a physical DVD from a video store, potential y representing a drop in revenue per movie watched at home. However, the cloud model of delivery al ows a service such as Netflix to serve mil ions of viewers across multiple countries, generating far greater revenue. The proliferation and sophistication of cloud services could be a boon to entrepreneurs and smal enterprises. Cloud platforms make it much easier and cheaper for smal businesses to pay for IT resources on a per-use basis, al owing them to scale their IT capacity up or down and build critical operational capabilities. The cloud could become a major force in making entrepreneurship more feasible in the coming decade. The growing role of the Internet as an enabler of economic growth makes the cloud a phenomenon that policy makers need to address. To make the benefits of cloud technology available to citizens, governments and policy makers should consider programs to build greater network capacity and create incentives for providing high-speed Internet service. In developing economies, policy makers can encourage the build-out of wireless broadband, which wil be the main way in which their citizens wil access the Internet.
68 Policy makers should take a thoughtful approach to regulations related to data ownership, security, privacy, and liability to remove uncertainty about cloud use. The law in most countries does not address these issues. In addition, data protection laws in many countries restrict the storage and transfer of several types of data outside their borders, which constrains the ability to take advantage of some of the benefits of cloud technology.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 69 5. Advanced robotics During the past few decades, industrial robots have taken on a variety of manufacturing tasks, usual y those that are difficult, dangerous, or impractical for humans—welding, spray-painting, or handling heavy materials, for example. Robotics is now seeing major advances that could make it practical to substitute machines for human labor in increasing numbers of manufacturing applications, in many service applications, and, importantly, in extremely valuable uses such as robotic surgery and human augmentation. Advances in artificial intel igence, machine vision, sensors, motors, and hydraulics—even in materials that mimic a sense of touch—are making this possible. Robots are not only becoming capable of taking on more delicate and intricate tasks, such as picking and packing or manipulating smal electronics parts, but they are also more adaptable and capable of operating in chaotic conditions and working alongside humans. At the same time, the cost of robots is declining. Advanced robotics promises a world with limited need for physical labor in which robot workers and robotic human augmentation could lead to massive increases in productivity and even extend human lives (see Box 7, “The promise of advanced robotics: Machines that end physical toil and improve lives”). Many goods and services could become cheaper and more abundant due to these advances. The physical y handicapped and the elderly could lead healthier and less-restricted lives using robotic prosthetics and “exoskeletons” that strap on like braces and assist in locomotion. We estimate that the application of advanced robotics across health care, manufacturing, and services could generate a potential economic impact of $1.7 tril ion to $4.5 tril ion per year by 2025, including more than $800 bil ion to $2.6 tril ion in value from health-care uses. This impact would result from saving and extending lives and transforming the way in which many products are built and many services are delivered. Advanced robotics also holds a great deal of promise for businesses and economies. Early adopters could gain important quality, cost, and speed advantages over competitors, while some companies could find that advanced robotics lowers the barriers for new competitors. Businesses in developing economies could be among the biggest buyers of robotics given the current rate of automation; however, these economies could be negatively impacted by fal ing demand for low-wage manual labor, upon which they rely for economic development. The ability of robots to take on a far wider range of jobs economical y could encourage global companies to move some production back to advanced economies. In advanced economies, some workers might find new job opportunities in developing, maintaining, or working with robots. At the same time, many jobs in advanced economies involving manual labor might be automated away, placing even more importance on educating and retraining workers for higher-skil jobs.
70 Box 7. Vision: Machines end physical toil and improve lives Imagine a world in which advanced robots expertly and inexpensively perform and augment most physical tasks. Imagine you are a manager in a manufacturing plant in 2035. At your plant, injuries are virtual y unheard- of. In fact, there are few people on the floor: a smal group of highly skil ed specialists oversee thousands of robots, interacting natural y with the robot workforce to produce goods with unprecedented speed and precision, 24 hours a day, 365 days a year. When a new product or design improvement is introduced, factory workers train robots to fol ow new routines, using simple touch-screen interfaces, demonstration, and even verbal commands. Most of your day is spent optimizing processes and flows and even assisting with product designs based on what you see on the factory floor and the data that your robots generate. During lunch, you swing by a local fast-food restaurant. You watch as your meal is prepared and cooked exactly the way you like it by a robot. Back at your desk, you see service robots making deliveries and cleaning the floors and windows. Outside, robots pick up trash and replace broken street lights. In a world of advanced robotics, surgeons are assisted by miniature robotic surgery systems, greatly reducing both the time necessary for procedures and their invasiveness. Recovery is more rapid as wel . People suffering from paralysis due to spinal injuries are able to walk again with the help of robotic exoskeletons directly connected to the nervous system. DEFINITION Traditional robots excel at tasks that require superhuman speed, strength, stamina, or precision in a control ed environment (robot welding or semiconductor fabrication, for example). They are bolted in place behind railings to prevent injuries to humans. They do exactly what they are programmed to do—and nothing more. But now, a new generation of more sophisticated robots is becoming commercial y available. These advanced robots have greater mobility, dexterity, flexibility, and adaptability, as wel as the ability to learn from and interact with humans, greatly expanding their range of potential applications. They have high-definition machine vision and advanced image recognition software that al ows them to position objects precisely for delicate operations and to discern a part in a pile. They are powered by sophisticated motors and actuators, al owing them to move faster and more precisely, and some are made from lighter, softer materials. The US Defense Advanced Research Projects Agency (DARPA) is even working on robots that can ful y automate the sewing of garments, using a process that tracks the movement of individual threads and precisely moves fabric to perform exact stitching.65 Advances in artificial intel igence, combined with improved sensors, are making it possible for robots to make complex judgments and learn how to execute tasks 65 Katie Drummond, “Clothes wil sew themselves in DARPA’s sweat-free sweatshops,” Wired, June 8, 2012.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 71 on their own, enabling them to manage wel in uncertain or fluid situations. By 2025 advanced robots could be capable of producing goods with higher quality and reliability by catching and correcting their own mistakes and those of other robots or humans. These robots can sense and quickly react to obstacles, other robots, or human coworkers, giving them greater “awareness” and making it possible for them to work more safely side-by-side with humans. Many advanced robots can also communicate with one another and work together on shared tasks. Some advanced robots are designed to be simple, smal , and inexpensive, while having the ability to be networked together and work in teams. These distributed, or “swarm,” robots could eventual y be used for dangerous tasks such as search and rescue operations. Final y, advances in interfaces, sensors (including sophisticated tactile sensors), and actuators, combined with improved materials and ergonomic designs, are furthering robotic surgery and dramatical y improving the quality and usefulness of human prosthetic devices. Ultraprecise surgical robots are making new forms of minimal y invasive surgery possible that can reduce postsurgical complications, enable faster recovery, and possibly reduce surgical death rates. Robotic prosthetics and exoskeletons are able to take precise directions and make increasingly accurate and delicate movements. New interfaces have been developed that can operate robotic limbs using smal electrical signals produced when muscles contract or using signals from nerve endings or even brain waves. The capabilities of these prosthetics may soon come to rival or exceed those of actual human limbs. These advances could eventual y include prosthetic hands with independently moving fingers and prosthetic body parts that mimic the sense of touch using a neural interface.66 These technological advances, combined with declining costs, are making entirely new uses for robots possible. For example, El Dulze, a Spanish food processor, now uses highly agile robots to gently pick up heads of lettuce from a conveyor belt, measure their density (rejecting heads that don’t meet company standards), and replace them on the belt, where other robots position the heads for a machine that removes their roots. The company says the robots are better than humans at assessing lettuce quality (the reject rate has fal en from 20 percent to 5 percent), and hygiene at the facility has also improved.67 POTENTIAL FOR ACCELERATION Adoption rates for advanced robots wil be determined by many factors, including labor market conditions. For example, in China, where wages and living standards are rising, workers are pressing for better working conditions, including relief from long hours of precise piecework that can lead to repetitive stress injuries. As education levels rise, fewer workers are wil ing to take such jobs. As a result, Foxconn, a contract manufacturer that employs 1.2 mil ion workers, is investing in robots to assemble products such as the Apple iPhone.68 According to the International Federation of Robotics (a major robotics industry group), China is expected to become the world’s largest consumer of industrial robots by 2014. Global manufacturing labor costs are $6 tril ion annual y today, so additional automation represents a huge opportunity. 66 Megan Scudel ari, “Missing touch,” The Scientist, September 1, 2012. 67 68 robots perform farmer’s work, case study of Fanuc Robotics Europe S.A., International Federation of Robotics, September 2012. 68 John Markoff, “Skil ed work, without the worker,” The New York Times, August 18, 2012.
72 Demographics wil also play a role in determining demand for advanced robotics. Robotic surgical systems and prosthetics could help meet the large and growing need (particularly in advanced, aging economies) to provide quality health care. And many manufacturers stil rely on legions of low-skil workers (often in developing countries) to do work that involves precise operations on irregular objects, such as bending tiny wires to assemble mobile phones or deboning chicken breasts; over the coming decade, many of these tasks could be automated. New applications for advanced robotics, particularly in services, are also emerging. Robots are now poised to take on dirty, dangerous, and labor-intensive service work, such as inspecting and cleaning underground pipes, cleaning office buildings, or col ecting trash. Domestic service robots are another expanding market. Though robotic vacuum cleaners have been around for years, sales of these and similar household robots are now growing rapidly, by about 15 to 20 percent annual y. Adoption could accelerate even further by 2025 as these machines become more capable and consumers consider the trade-offs between buying robots, sacrificing leisure time, or hiring professional cleaners or gardeners to perform these tasks. Advanced robots are also of great interest to military planners, who see opportunities to both automate combat (similar to remotely piloted drone aircraft) and support human troops. DARPA is investing in a range of advanced robotics programs, from a ful robotics “chal enge” (similar to the DARPA Grand Chal enge that pioneered self-driving cars) to four-legged robots for carrying supplies, robotic exoskeletons and suits to strengthen and protect troops, and advanced prosthetic limbs to help injured soldiers. This type of military investment could greatly speed further advancement. Robot prices are dropping, placing them within reach of more users. Industrial robots with features such as machine vision and high-precision dexterity typical y cost $100,000 to $150,000. By 2025, it is possible that very advanced robots with a high level of machine intel igence and other capabilities could be available for $50,000 to $75,000 or less. In recent decades, robot prices have fal en about 10 percent annual y (adjusted for quality improvements) and may decline at a similar or faster rate through 2025.69 Accelerated price declines could be made possible by scale efficiencies in robot production (due in large part to rising demand by Chinese and other Asian manufacturers), the decreasing cost of advanced sensors (partly driven by demand for inexpensive sensors in smartphones and tablets), and by the rapidly increasing performance of computers and software. Some entrepreneurs are focusing on developing inexpensive general purpose robots that can be easily trained to do simple tasks (see Box 8, “Your new coworker, Baxter”). The rate at which robots could proliferate is a subject of intense debate. According to the International Federation of Robotics, industrial robot sales reached a record 166,000 units in 2011, a 40 percent jump over 2010; sales in China grew by more than 50 percent in 2011. Since 1995 global sales have grown by 6.7 percent per year on average. It is possible that there could be even faster growth ahead if Baxter and other low-priced, general-purpose models can drive rapid adoption in simple manufacturing and service work. At the same time, instal ations of advanced industrial robots could accelerate beyond historic rates if 69 World robotics 2012, International Federation of Robotics, August 30, 2012.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 73 robotics technology continues to accelerate. Adoption scenarios wil depend both on improvements in capability and price and receptivity to automation; in addition, significant organizational and societal barriers may stand in the way. Box 8. Your new coworker, Baxter To make robots useful in low-end manufacturing, they not only have to be priced attractively, but they also need to fit into the workplace. They can’t take up too much space, they have to work wel and safely with humans, and they have to be easy to program. These were some of the goals for “Baxter,” a $22,000 general-purpose robot developed by startup company Rethink Robotics. Another goal was to put a friendly face on robots—literal y. Baxter features an LCD display screen mounted on a “neck” above its body. The screen shows a pair of eyes that take on different expressions depending on the situation. The eyes fol ow what the robot’s two arms are doing, as a human worker would. While Baxter’s functionality is somewhat limited—it is best at performing simple operations such as picking up objects, moving them, and putting them down—it makes up for these limitations with superior adaptability and modularity created by the ability to instal different standard attachments on its arms. When the robot is first instal ed or needs a new routine, it “learns” without the need for programming. A human simply guides the robot arms through the motions that wil be needed for the task, which Baxter memorizes. It even nods its “head” to indicate that it has understood its new instructions. POTENTIAL ECONOMIC IMPACT We estimate that by 2025 advanced robotics could have a worldwide economic impact of $1.7 tril ion to $4.5 tril ion annual y across the applications we have sized (Exhibit 7). Much of this impact—$800 bil ion to $2.6 tril ion—could come from improving and extending people’s lives. An additional $700 bil ion to $1.4 tril ion could arise from automating manufacturing and commercial service tasks. We estimate that the use of advanced robots for industrial and service tasks could take on work in 2025 that could be equivalent to the output of 40 mil ion to 75 mil ion ful -time equivalents (FTEs). This could potential y have annual economic impact of $600 bil ion to $1.2 tril ion in developed countries and $100 bil ion to $200 bil ion in developing economies. Final y, $200 bil ion to $500 bil ion in impact could arise from the use of time-saving household service robots.
Exhibit 7 Sized applications of advanced robotics could have direct economic impact of $1.7 trillion to $4.5 trillion per year in 2025 Potential economic impact of sized Sized applications in 2025 Estimated scope Estimated potential Potential productivity applications $ tril ion, annual y in 2025 reach in 2025 or value gains in 2025 Robotic 50 mil ion amputees and 5–10% of amputees $240,000–390,000 per 0.6– human people with impaired and people with person for extended/ 2.0 augmentation mobility in advanced impaired mobility in improved quality of economies advanced economies life1
355 mil ion applicable 30–60 mil ion FTEs of 75% potential Industrial 0.6– industrial workers work potential y improvement in
robots 1.2 automatable across productivity per unit of
key job types work automated
Surgical 0.2– 200 mil ion major 5–15% of major 60,000–180,000 lives robots 0.6 surgeries in countries with surgeries in countries saved per year developed health care with developed health- 50% reduction in sick care systems and inpatient days Personal 0.2– and home 90–115 bil ion hours spent 25–50% of households 20–50 bil ion hours 0.5 robots on tasks such as cleaning in advanced saved per year and lawn care per year in economies $10 value per hour of advanced economies time saved Commercial 0.1– 130 mil ion applicable 10–15 mil ion FTEs of 35–55% potential service robots 0.2 service workers work potential y improvement in automatable across productivity per unit of Other key job types work automated potential applications (not sized) Sum of sized potential 1.7– economic 4.5 impacts 1 Using QALY (quality-adjusted life years) estimates. NOTE: Estimates of potential economic impact are for some applications only and are not comprehensive estimates of total potential impact. Estimates include consumer surplus and cannot be related to potential company revenue, market size, or GDP impact. We do not size possible surplus shifts among companies and industries, or between companies and consumers. These estimates are not risk- or probability-adjusted. Numbers may not sum due to rounding. SOURCE: McKinsey Global Institute analysis Health care We estimated the potential economic impact of robotic surgery and robotic prosthetics to be as much as $800 bil ion to $2.6 tril ion annual y by 2025, based on saving lives and improving quality of life. For estimating the potential economic impact of robotics for human augmentation, we considered potential uses of robotic prosthetics and exoskeletons.70 By 2025 there could be more than 50 mil ion people with impaired mobility in the developed world, including amputees and elderly people, for whom robotic devices could restore mobility, improve quality of life, and increase lifespan. It is possible that 5 to 10 percent of these people could have access to robotic augmentation by 2025 given the current penetration of alternatives such as traditional prosthetics and motorized wheelchairs. Studies indicate that impaired mobility contributes significantly to reduced life expectancy due to increased health risks such as injury and osteoporosis.71 If it were possible to extend life by one to two years for each disabled person and provide a 20 to 30 percent improvement in quality of life over eight years using 70 Robotic mechanisms that can be worn 7 by physi caly handicapped people to help move limbs (or even entire bodies). 71 For more on the effects of disabilities on life expectancy, see R. Thomas and M. Barnes, “Life expectancy for people with disabilities,” NeuroRehabilitation, May 2007.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 75 robotic assistance (assuming substantial restoration of normal function) the result could be a potential impact of $240,000 to $390,000 per person, using a quality- adjusted life year (QALY) approach. If these results can be achieved, robotics for human augmentation could lead to a potential economic impact of $600 bil ion to $2.0 tril ion per year by 2025, much of which could be consumer surplus accruing to the users of these robotic devices. As the technology for robotic surgery improves, it could have the potential to prevent deaths and significantly reduce both in-patient care time and missed work days. Robotic surgical “platforms” are already being used for minimal y invasive procedures such as laparoscopic surgery. It is possible that with advances in robotic technology, by 2025 robotic surgery could be widely used for these and other procedures. Approximately 200 mil ion major surgeries could be performed every year in countries with developed health-care systems in 2025.72 Currently, about 3 percent of al major surgeries result in death, but it is possible that by 2025, advanced robotic surgical systems could help reduce these deaths substantial y, perhaps by as much as 20 percent, by reducing common complications such as bleeding or internal bruising. This improvement in outcomes could be enabled by more flexible surgical robots with a greater range of motion that could perform more types of operations, or from new features such as AI-assisted autocorrect systems that could warn surgeons when they are about to cut the wrong tissue or apply too much pressure. Declining costs in robotic surgery systems could al ow more hospitals and surgeons to use the technology, potential y increasing the performance of many surgeons. We estimate that if 5 to 15 percent of al major surgeries in countries with developed health-care systems could be performed with the assistance of robots by 2025, it could result in 60,000 to 180,000 lives saved each year. Robot-assisted surgery could also cut in-patient stays and sick days associated with surgery by 50 percent. If these results can be achieved, we estimate that robotic surgery could have an economic impact of $200 bil ion to $600 bil ion per year by 2025.73 Industrial robots For industrial robots, we analyzed data regarding job tasks, occupations, and distribution across countries.74 We then considered which tasks could be ful y or partial y automated economical y by advanced robots by 2025, assuming a high level of robot performance and continued reductions in cost. In developed countries, across occupations such as manufacturing, packing, construction, maintenance, and agriculture, we estimate that 15 to 25 percent of industrial worker tasks could be automated cost-effectively (based on estimated 2025 wage rates) by 2025. We estimate that in developing countries, on average, 5 to 15 percent of manufacturing worker tasks could be automated across relevant occupations by 2025. 72 Based on analysis by Thomas G. Weiser et al., “An estimation of the global volume of surgery: A modeling strategy based on available data,” Lancet, volume 372, number 9633, July 12, 2008. 73 We use a quality of adjusted life year (QALY) of $100,000 and assume that surgical patients avoiding death are restored to a normal life expectancy. 74 Analysis of job occupations and tasks is based on data from a variety of sources, including labor and wage data from the Economist Intel igence Unit, International Labour Organisation, IHS Global Insight, Eurostat, and various national labor bureaus.
76 We calculated the potential cost savings using the estimated annual cost of advanced robots compared with the annual employment cost of an equivalent number of workers. This yields a potential economic impact of $600 bil ion to $1.2 tril ion per year by 2025. This would imply a substantial increase in the number of industrial robots instal ed global y by 2025, by about 15 mil ion to 25 mil ion robots, requiring investments totaling about $900 bil ion to $1.2 tril ion. Realizing al of this potential impact would therefore imply 25 to 30 percent average annual growth in robot sales, significantly higher than the average growth rate over the past two decades, but lower than the growth rate in 2010 and 2011. Service robots Service robots fal into two categories: those used in commercial settings and personal robots. For personal and household service robots, we focused on the potential to automate cleaning and domestic tasks such as vacuuming, mopping, lawn mowing, and gutter cleaning. The use of advanced robots for these types of tasks has significant potential given the current trajectory of technology improvement, the relatively low cost of the robots required, and the already increasing rate of adoption. Sales of household robots used largely for the above-mentioned tasks are already growing by about 20 percent annual y. To estimate the potential impact of household robots, we considered the amount of time spent on relevant cleaning and domestic tasks, focusing on the developed world, where significant adoption is most likely. Based on US and European labor studies, we estimate that 90 bil ion to 115 bil ion hours per year are spent performing relevant household tasks in the developed world.75 If 25 to 50 percent of people in the developed world were to adopt the use of these robots by 2025, $200 bil ion to $500 bil ion worth of time savings could be realized. We believe this level of adoption is possible given the rapid advances in low-cost robotics technology, the relatively limited sophistication of the robots required for these applications, and the demonstrated wil ingness of many consumers to pay for household time-saving devices. For commercial service robots, we analyzed data on job tasks, occupations, and distribution across countries.76 We then considered which tasks could be ful y or partial y automated economical y by advanced robots by 2025, assuming a high level of robot performance and continued reductions in cost. We estimate that in developed economies, across occupations such as food preparation, health care, commercial cleaning, and elder care, as much as 7 to 12 percent of commercial service worker tasks could be automated cost-effectively by 2025. For example, nurses spend up to 20 percent of their shift time wheeling equipment and carts from one location to another or waiting for a cart to arrive. So-cal ed courier robots (self-guided, motorized carts) can take on these tasks. We estimate that in developing countries, 4 to 8 percent of commercial service worker tasks could be automated across relevant occupations by 2025. To achieve this, we estimate that 2.5 mil ion to eight mil ion advanced robots would be necessary, requiring an estimated investment of $200 bil ion to $400 bil ion global y by 2025. 75 Estimates based on data from Rachel Krantz-Kent, “Measuring time spent in unpaid household work: Results from the American time use survey,” Monthly Labor Review, volume 132, number 7, July 2009. 76 Analysis of job occupations and tasks based on data from a variety of sources, including labor and wage data from the Economist Intel igence Unit, International Labour Organisation, IHS Global Insight, Eurostat, and various national labor bureaus.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 77 BARRIERS AND ENABLERS There are several important barriers that could limit adoption of advanced robotics by 2025. First, although costs are declining, most industrial and many commercial service robots remain expensive, costing tens or hundreds of thousands of dol ars per robot. Surgical robots often cost more than $1 mil ion (although these costs could come down and few of these might be needed compared with industrial and service robots). Large-scale adoption of industrial and service robots could require investments of perhaps $1.1 tril ion to $1.6 tril ion by 2025. Before making these investments, companies would likely require strong evidence of positive returns on investment, and establishing a clear track record of performance could take years. And once robots are purchased and instal ed, it can stil take time to redesign processes and flows to ful y take advantage of their capabilities. Surgical robots have already seen significant growth in adoption, but additional trials and data wil be required to ful y demonstrate their benefits, particularly given concerns about health-care costs in the United States and other advanced economies. There are also questions about whether robotical y assisted surgery has demonstrated significantly better performance than nonrobotic minimal y invasive surgery techniques, which are less costly. While the capabilities and performance of the technology could improve significantly by 2025, adoption may be constrained until definitive proof of results is available. The talent required to operate and maintain advanced robots is an important enabler that wil be required to ful y capture their potential. Some advanced robots could be designed to be very user-friendly and able to work natural y side-by-side with humans, but advanced robots may stil require a high level of expertise to maintain their hardware and software. Final y, the potential effect of advanced robots on employment could generate social and political resistance, particularly if robots are perceived as destroying more jobs than they create. Although the productivity improvements that advanced robots would create would drive growth in the economy, the workers who would be displaced might not be easily re-employed. Policies discouraging adoption of advanced robots—for example, by protecting manual worker jobs or levying taxes on robots—could limit their potential economic impact. Policy makers wil face difficult questions regarding legal liability, such as determining who is at fault when service or household robots contribute to accidents or injuries. IMPLICATIONS Over the coming decade, advanced robotics could deliver tremendous value for robot creators, health-care providers, manufacturers, service providers, entrepreneurs, consumers, and societies. For many businesses, advanced robotics promises significantly reduced labor costs, greater flexibility, and reduced time to deliver products to the marketplace. Business leaders should look for opportunities to leverage growing technology capabilities to help automate difficult, labor-intensive, and dangerous tasks in ways that are simple, user-friendly, and cost-effective, whether for treating patients or automating manual work.
78 For hospitals and health-care providers, advanced robotics could ultimately offer substantial improvements in patient care and outcomes. As a result, providers of robotic systems and supporting tools or services could see large growth opportunities over the coming decade. Makers of robotic surgical systems could see strong demand for increasingly advanced systems, but also feel pressure to minimize costs and clearly demonstrate improved outcomes for patients. Makers of robotic prosthetics and exoskeletons could experience similarly high demand, and may want to look for ways to reach disabled and elderly people around the world, even in less-developed regions. Manufacturing and service companies with large workforces could benefit from reduced costs, reduced injuries, and lower overhead, as wel as reducing payrol s in human resources, labor relations, and factory supervisory roles. Factories might no longer need to be located near sources of low-cost labor, al owing them to be located closer to final assembly and consumers, simplifying supply chains and reducing warehousing and transportation costs. However, business leaders wil face chal enges in capturing the ful productivity and quality improvements that could be afforded by advanced robots. Advanced robotics requires substantial capital investments, and businesses wil need clear evidence of positive return on investment. Reconfiguring manufacturing processes, service delivery channels, and supply chains is difficult and time consuming. Training employees to work effectively alongside robots is also no smal task. To maximize value capture and stay ahead of the curve, businesses should continual y experiment with advanced robotics and additional automation, identify promising technologies, rethink business processes, and develop in- house talent. They should also consider how their supply chains could be redesigned to leverage automation, and how additional speed to market, flexibility, and quality could help differentiate their offerings from those of competitors. For some entrepreneurs, decreasing robot cost and increasing capabilities could make entirely new business models possible or decrease barriers to entry in the manufacturing and service industries. Robotical y enabled production facilities, fast-food restaurants, self-service laundries, and medical clinics might offer superior efficiency and quality and could scale quickly. Established manufacturers may need to accelerate automation to meet the competition while investing in innovative product development or superior service quality to better differentiate their offerings. For societies and policy makers, the prospect of increasingly capable robots brings potential benefits: growing national productivity, higher-quality goods, safer surgeries, and better quality of life for the elderly and disabled. But it also poses new chal enges in employment, education, and skil training. In some cases, access to advanced robotics could cause companies to repatriate manufacturing operations from low-wage offshore locations. And the spread of robotics could create new high-skil employment opportunities. But the larger effect could be to redefine or eliminate jobs. By 2025, tens of mil ions of jobs in both developing and advanced economies could be affected. Many of these employees could require economic assistance and retraining. Part of the solution wil be to place even more emphasis on educating workers in high-skil , high-value fields such as math, science, and engineering.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 79 6. Autonomous and near-autonomous vehicles For years, almost al commercial aircraft have had the ability to operate on autopilot. Onboard computers can manage most aspects of flying, including even aspects of takeoff and landing. The tankers and cargo ships that transport most of the goods for the global economy are highly automated, al owing them to operate with very smal crews. Now partly or completely self-driving cars and trucks are also becoming a reality, enabling a potential revolution in ground transportation that could, if regulations al ow, be wel under way by 2025. Autonomous vehicles offer several potential benefits, including reducing deaths from motor vehicle crashes and reducing CO emissions. With computers 2 control ing acceleration, braking, and steering, tightly spaced cars and trucks can safely travel at higher speeds; when one vehicle in line brakes or accelerates, they al do. Since most driving accidents are caused by human error, removing drivers could actual y increase traffic safety and reduce deaths, injuries, and property losses. Convoys of trucks could speed down the highway with no driver needed (or just one driver in the lead truck), with as little as one foot of space between them.77 Roadways could accommodate more vehicles without expansion, and acceleration and braking could be optimized to reduce fuel consumption and CO 2 emissions. In addition, closely spaced vehicles have much lower aerodynamic drag, which further reduces fuel consumption. Drivers could be free to use their drive time to work, relax, or socialize while being transported. If regulators approve autonomous driving and the public accepts the concept, the benefits provided by improved safety, time savings, productivity increases, and lower fuel consumption and emissions could have a total economic impact of $200 bil ion to $1.9 tril ion per year by 2025. Technology is not likely to be the biggest hurdle in realizing these benefits. In fact, after 20 years of work on advanced machine vision systems, artificial intel igence, and sensors, the technology to build autonomous vehicles is within reach—as a growing number of successful experimental vehicles have demonstrated. What is more likely to slow adoption is establishing the necessary regulatory frameworks and winning public support. In order to realize other benefits (which we have not sized), such as reduced congestion, infrastructure investments would be needed to create special lanes and instal sensors to control traffic flow on major arteries. And there wil be legal and ethical questions to address, such as who bears responsibility when an autonomous vehicle causes an accident and how to program a computer to make life-and-death decisions (such as weighing whether to swerve to avoid a pedestrian against the chance of injuring passengers). Nevertheless, autonomous vehicles are coming, in fact, some autonomous features, such as self-parking systems, are already available in production vehicles. While the economic impact driven by this technology could be quite large, it may take many years to ful y materialize. 77 Brian Dumaine, “The driverless revolution rol s on,” Fortune, November 12, 2012.
80 DEFINITION An autonomous vehicle is one that can maneuver with reduced or no human intervention. In this report, we focus on autonomous cars and trucks, which we believe have the greatest potential for significant economic impact by 2025. Other forms of autonomous vehicles—such as crop-spraying drone aircraft, self-guided forklift trucks, and law enforcement drones—may also become widely used, but we believe they have more limited applications and less incremental impact within our time frame. Also, we have chosen not to include estimates of the potential value of autonomous military vehicles and drones in the context of this report. Machine vision is a key enabling technology for autonomous vehicles. Using cameras and other sensors, a computer constantly monitors the road and the surrounding environment, acquiring an image and then extracting relevant information (such as stop signs or objects in its path) on which to base actions. Advances in machine vision include 3D cameras that gather additional information regarding distances that two-dimensional cameras cannot provide. Pattern recognition software, including optical character recognition programs, can interpret symbols, numbers, and the edges of objects in an image. LIDAR (laser- imaging detection and ranging), which is similar to radar but uses laser light bounced off of objects rather than radio signals to measure distance, is also being used by autonomous vehicles, along with advanced GPS (global positioning system) technologies and spatial data. When combined with sensor data, this information enables autonomous vehicles to pinpoint their current locations, fol ow the road, and navigate to their destinations. Input signals from machine vision and sensors are integrated with stored spatial data by artificial-intel igence software to decide how the vehicle should operate based on traffic rules (for example, obeying speed limits and yields signs) and knowledge of exceptions (such as stopping when the light is green if a pedestrian is in the intersection). Control engineering software does the “driving,” giving instructions to the actuators that perform the task needed for the desired action, such as accelerating, braking, or turning. With these capabilities, a ful y autonomous vehicle can navigate to a specified destination, moving safely among other vehicles, obstacles, and pedestrians. These vehicles’ computers can also optimize fuel economy by accelerating and braking smoothly, remaining within the speed limit, and never taking a wrong turn. Google has demonstrated these capabilities with a Toyota Prius that has been equipped with computers, sensors, actuators, and other technology; this vehicle has been driven for 300,000 miles with only one accident (which was human- caused).78 Partly autonomous driving features—including steering assistance (maintaining the car’s position between lane markers), braking and accelerating to maintain distance from vehicles ahead, and automatic braking when obstacles appear ahead—are already being offered or wil soon be offered on production vehicles. In the next decade, we can expect autonomous driving to be offered as an option on new automobiles, initial y on high-end models and later on mid-priced vehicles. Eventual y, autonomous driving could give rise to new kinds of vehicles. These might include driverless passenger vehicles (which would not require a driver to 78 Frederic Lardinois, “Google’s self-driving cars complete 300K miles without accident, deemed ready for commuting,” TechCrunch, August 7, 2012.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 81 sit behind the wheel) that could be configured to maximize work space or even provide beds for passengers; new concepts involving car sharing in which a car could arrive or leave and park wherever and whenever needed; or new public transportation vehicles that would al ow for greater flexibility and personalization. POTENTIAL FOR ACCELERATION Autonomous driving would herald a new era for automobiles. During the past century, the automobile was a breakthrough general-purpose technology that provided the means for getting workers to their jobs and consumers and goods to markets. Automobiles enabled economic development and higher living standards around the world. Trucks have spread economic development and markets to locations that railroads, rivers, and canals have never been able to reach. However, these machines have also caused pol ution, soaring demand for fossil fuels, and congestion and related productivity losses, as wel as death and injury. The average American car owner spends 750 hours per year driving—the equivalent of four months of work days—while the average European spends about 300 hours.79 More than one mil ion people are kil ed in traffic accidents every year around the world, and it is estimated that 70 to 90 percent of all automobile accidents are caused by human behavior.80 The majority of this waste and destruction could be avoided by using autonomous vehicles. Around the world, autonomous vehicles have the potential to improve the economics of trucking. Self-driving trucks that transport goods long distances could fit easily into intermodal transportation and logistics systems. Trucks moving in convoys could transport goods on major arteries, then transfer their cargos at regional distribution centers, from which other vehicles could take the cargo to its final destinations. On their long hauls, autonomous trucks would not have to stop for their drivers to sleep or eat. The technology for autonomous vehicles has evolved at lightning speed. In 2004 DARPA sponsored a $1 mil ion prize for driverless vehicles to navigate a course in the Mojave Desert cal ed the “DARPA Grand Chal enge.” No teams finished the race. One year later, in 2005, five cars successful y crossed the finish line. In 2007 a more urban version of the race was held incorporating street signs, obstacles, and traffic; six teams finished.81 Today, Google’s self-driving cars are driving on city streets and freeways (with a human driver behind the wheel as backup in case the system has a problem) in the US states of California and Nevada. Google has announced that it expects to have a commercial y available version of its technology ready in three to five years. The self-driving technology that is being tested today would add thousands of dol ars to the price of a car, but the cost of these systems is expected to drop. For example, researchers at Oxford University are aiming to develop an autonomous system that would cost as little as $150.82 Even though the regulatory framework for autonomous vehicles is not yet in place (California and Nevada are permitting testing on public roads), major automakers 79 Mobility choices: Consumers at the wheel, McKinsey & Company survey, June 2012. 80 Hans von Holst, ed., Transportation, traffic safety and health: The new mobility, Springer Publishing, 1997. 81 Darpagrandchal enge.com. 82 “Researchers testing frugal autonomous car systems, aim for $150 price tag,” Engadget. com; R. W. Wal , J. Bennett, G. Eis, “Creating a low-cost autonomous vehicle,” presented at Industrial Electronics Society annual conference in Sevil a, Spain, November 5–8, 2002.
82 are moving ahead with development. General Motors, Toyota, Mercedes-Benz, Audi, BMW, and Volvo are testing their own autonomous systems. Audi is testing what it cal s a “piloted” car that can handle starts and stops in heavy traffic and park itself.83 The driver can take over control of the vehicle at any time and is expected to always monitor the vehicle’s driving. Cadil ac has demonstrated an enhanced cruise-control system that not only controls speed but also provides steering assistance on highways. The 2014 Mercedes-Benz S-class is arriving soon with a range of advanced, though not ful , autonomous driving capabilities standard, including keeping in a lane and maintaining speed and distance from other cars. These capabilities will only be accessible under certain driving conditions and drivers wil need to keep their hands on the wheel, but its developers claim the onboard vision system can keep the car in its lane and maintain safe distances to other vehicles at speeds up to 124 miles per hour. The driver needs to step in when the car changes driving environments (such as when exiting a freeway). In congested traffic, the car will be able to track the environment and understand when to accelerate and when to brake.84 Japan’s New Energy and Industrial Technology Development Organization, a research organization, has successful y tested an autonomous trucking system in which a single driver leads three other trucks that are equipped with roof- mounted radar systems, traveling at 50 miles per hour, spaced about four meters apart. On-site autonomous vehicles are also being tested by mining giant Rio Tinto. The company has used 150 partly autonomous trucks in Australian mining operations. The trucks fol ow a predefined route and load and unload material without an operator. POTENTIAL ECONOMIC IMPACT The potential economic impact of autonomous cars and trucks could be $200 bil ion to $1.9 tril ion per year by 2025 (Exhibit 8). This is an indicative value estimation of the benefits of the technology that could be realized if autonomous vehicles are al owed by regulations and adopted by consumers. The largest impact would come from freeing up time for drivers, increased road safety, and the reduced cost of operation of vehicles. We estimate that 30,000 to 150,000 lives could be saved per year in 2025 if this technology is adopted and that CO emissions could be reduced by as much as 300 mil ion tons per year. That 2 amount is equivalent to 50 percent of CO emissions from current commercial 2 aviation.85 Self-driving cars could have a potential economic impact of $100 bil ion to $1.4 tril ion per year in 2025. This assumes that 75 to 90 percent of cars sold from 2017 to 2020 in the high-end auto segment, as wel as 20 to 30 percent of midpriced cars, could have self-driving capability. That would translate into approximately 10 to 20 percent of the 1.2 bil ion private cars projected to be on the road in 2025 having the ability to self-drive in at least half of all traffic situations. 83 Angus MacKenzie, “The future is here: Piloted driving in Audi’s autonomous A6,” MotorTrend, January 2013. 84 Andrew English, “New car tech: 2014 Mercedes-Benz S-class,” Road & Track, November 2012. 85 Introduction of aviation into the European Union emissions trading scheme, Econometrica Press summary paper, May 2009.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 83
Exhibit 8 Sized applications of autonomous and near-autonomous vehicles could have direct economic impact of $200 billion to $1.9 trillion per year in 2025 Potential economic impact of sized Sized applications in 2025 Estimated scope Estimated potential reach Potential productivity applications $ tril ion, annual y in 2025 in 2025 or value gains in 2025 900 mil ion new 5–20% of al driving $2–8 per hour in cars produced in or autonomous value of time saved Autonomous 0.1– after 2018 – 20–30% of cars sold from 70–90% fewer cars 1.4 500 hours per year 2017–20 with potential to accidents spent in car by be autonomous 15–20% gain in fuel
average owner – 50–100% driving time efficiency spent under ful computer
Autonomous 0.1– control
trucks 0.5 24 mil ion trucks 10–30% of new trucks with 70–90% fewer produced in 2018 or autonomous driving accidents later capabilities 10–40% greater fuel Other 50% driven by human efficiency potential drivers 1–2 drivers per 10 applications trucks (for monitoring) (not sized) Potential applications not sized include commercial drones, military drones, Sum of sized and/or autonomous and near-autonomous submersible vehicles for applications potential 0.2– such as fossil fuels exploration economic 1.9 impacts NOTE: Estimates of potential economic impact are for some applications only and are not comprehensive estimates of total potential impact. Estimates include consumer surplus and cannot be related to potential company revenue, market size, or GDP impact. We do not size possible surplus shifts among companies and industries, or between companies and consumers. These estimates are not risk- or probability-adjusted. Numbers may not sum due to rounding. SOURCE: McKinsey Global Institute analysis Assigning a value of $2 to $8 per hour of driving time that could be regained (depending on factors such as degree of car autonomy and potential legislation requiring the driver to pay attention to the road), we estimate that $100 bil ion to $1.0 tril ion in saved time could be freed up per year by 2025.86 We calculate the second-largest source of savings—reducing vehicular deaths—by estimating that self-driving technology can reduce road accidents by 5 to 20 percent overal . This assumes that autonomous cars wil not be subject to the type of accidents that are caused by human behavior. This could reduce annual automobile deaths by 30,000 to 140,000 a year global y in 2025.87 Fuel savings of 15 to 20 percent could be possible through the use of self-driving technology. Autonomous vehicles can be programmed to eliminate the 10 to 15 percent of fuel waste that occurs with rapid acceleration, speeding, and speed variation. Also, because sensors on each car respond instantly to the actions of the cars they fol ow, autonomous cars could travel more closely together, reducing air resistance and improving fuel efficiency by 15 to 20 percent when vehicles travel in convoy style.88 The consequent CO emission reduction could be 2 approximately 20 mil ion to 100 mil ion tons per year. 86 Based on US Department of Transportation estimates of the value of transportation time ($7 to $19 per hour) for different countries. The US Bureau of Labor Statistics estimates the value of work time at $16 per hour. We have chosen $2 to $8 per hour, because the ful value of the time cannot be recaptured while stil in t 8 he car. 87 We use a value per ful QALY (quality of adjusted life year) value of $100,000 for advanced economies and $50,000 in the developing world, with discount rate of 4 percent per year. 88 Kevin Bul is, “How vehicle automation wil cut fuel consumption,” MIT Technology Review, October 24, 2011.
84 Assuming that autonomous trucks could travel long distances in tightly spaced caravans and that drivers would not be needed in most vehicles, self-driven trucking could have a potential economic impact of $100 bil ion to $500 bil ion per year in 2025. We think it is possible that between 2017 and 2025, 10 to 30 percent of trucks sold could be at least partial y autonomous. Self-driving trucks potential y could be used for long-distance highway driving. However, over shorter distances, or for local deliveries, a driver would stil be needed. But even local delivery trucking could benefit from autonomous driving features that improve fuel efficiency and safety. We assume that half of autonomous trucks could stil have a driver (to provide service or make deliveries, for example); for convoys of ful y self-driving trucks, there might be one or two drivers for every ten trucks. If adoption of autonomous trucks occurs at the rates calculated here, we estimate economic impact of higher driver productivity could be $100 bil ion to $300 bil ion per year in 2025. Autonomous trucks could also prevent 2,000 to 10,000 traffic deaths per year in 2025 (as with passenger cars, 70 to 90 percent of trucking accidents are caused by human behavior, such as fal ing asleep at the wheel). Based on evolving technology, it is possible that autonomous trucks could be spaced less than three feet apart while driving, reducing fuel consumption by 15 to 20 percent by sharply reducing air resistance. Combined with speed control optimized for fuel efficiency, we estimate that autonomous trucks can use 10 to 40 percent less fuel than non-autonomous trucks. BARRIERS AND ENABLERS Governments wil have a central role in determining whether the potential value of autonomous vehicles wil be realized. Government efforts to encourage the development and ultimate adoption of autonomous cars and trucks could greatly speed their impact by helping to overcome concerns about technology, safety, liability, and legal responsibilities. Laws regarding autonomous driving wil be a critical enabler. If governments establish regulations early on that let autonomous vehicles travel on public roads, it wil provide a foundation on which to build new approaches to ground transportation. However, if regulations forbid drivers to take their hands off the wheel under any circumstances, the consumer surplus derived from saved driving time wil not materialize. If policy makers decide that the benefits of autonomous vehicles constitute a valuable public good, they can maximize those benefits by investing in intel igent road infrastructure systems that would make hands-off driving safer. Intel igent roads would have embedded sensors to provide precise positioning information and unambiguous input about speed limits. Other elements would include sensors at intersections to tel the vehicle if it is safe to proceed or if a traffic light is red or green, for example. In this way, the smart road takes on many of the complex tasks involved in guiding a vehicle safely. We do not believe that it is likely that extensive investments in intel igent roads wil be made by 2025. However, if the adoption of autonomous vehicles accelerates, the pressure to invest in autonomous-vehicle lanes and road sensors could rise as motorists seek to maximize the benefits of autonomous vehicles.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 85 Despite the technological progress that is now being seen in experimental vehicles, these systems stil require a great deal of improvement. Continuing work is required on vision, pattern recognition, and artificial-intel igence technologies to account for unexpected vagaries in infrastructure (for example, what to do when lane marker lines are obscured or traffic is rerouted around work crews). IMPLICATIONS Autonomous vehicles have significant potential to transform ground transportation, creating many opportunities for businesses and addressing many societal needs. They also have the potential to affect everyone who uses a car, al industries related to cars and trucks, and intermodal logistics systems. Autonomous vehicles could create great opportunities for new players in the automotive industry, including new types of competitors from technology and IT-related industries. In addition to providing vehicles with the input devices and onboard intel igence to maneuver independently, companies may find new business models that capitalize on the free time of drivers-turned-passengers (entertainment services or worker productivity tools designed for use in cars, for example). There is the possibility of a few players dominating the autonomous vehicle market, and these early entrants could create standards for operating systems and programming interfaces that might influence regulatory requirements. Early entrants into the autonomous-vehicle “ecosystem” could therefore benefit. As ful y autonomous vehicles become more accepted, there may also be opportunities for new types of driverless passenger cars (such as models designed for ride-sharing and carpooling that have no driver’s seat). This could provide an opening for new vehicle manufacturers or players in other industries to enter the market; it could also reduce rates of private car ownership. The success of autonomous cars and trucks could change the auto insurance industry. A significant reduction in traffic accidents and insurance claims could lead to a corresponding reduction in premiums. Individuals wil also have more complex risk profiles as they gain the ability to operate both self-driving and traditional cars. Eventual y, this could even drive a shift from traditional personal auto insurance to product liability insurance. However, some degree of personal liability will likely remain. Self-driving vehicles could have very disruptive effects on the trucking industry. In the United States, there are around 3.5 mil ion truck drivers. Demand for long- haul truck drivers would decline significantly, relegating truck driving to final-mile transportation and delivery. Companies should begin to work with employees to manage this change before this transition. The job of truck driver may come to involve more customer service, for example. Other driving jobs, such as taxi drivers and bus drivers, could also be at risk in the long term. Policy makers wil need to devise thoughtful supporting regulations for autonomous vehicles. Car manufacturers might not risk taking on the liability of driverless vehicles until regulators establish appropriate rules; for example, systems such as those that wil be available on the forthcoming Mercedes-Benz S-class are designed to disengage if the driver takes even one hand off the wheel above certain speeds.
86 Policy makers should also plan ways to maximize the value of autonomous vehicles to the economy and society. This should include long-term infrastructure planning that takes autonomous vehicles into account (for example, the creation of dedicated lanes for autonomous vehicles and sensor-embedded roads) and regulations that balance safety with the adoption of valuable technology. Autonomous vehicles wil also present legitimate security concerns. Like any computer system, a car’s autonomous guidance system could be hacked, with potential y disastrous results. Robust cyber security systems wil need to be in place before this technology hits the road.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 87 7. Next-generation genomics The science of genomics is at the beginning of a new era of innovation. The rapidly declining cost of gene sequencing is making huge amounts of genetic data available, and the ful power of information technology is being applied to vastly speed up the process of analyzing this data to discover how genes determine traits or mutate to cause disease. Armed with this information, scientists and companies are developing new techniques to directly write DNA and insert it into cel s, building custom organisms and developing new drugs to treat cancer and other diseases. Over the coming decade, next-generation genomics technology could power rapid acceleration in the field of biology and further alter health care. Desktop gene-sequencing machines are not far off, potential y making gene sequencing part of every doctor’s diagnostic routine. Longer term, these advances could lead to radical new possibilities, including ful y tailoring or enhancing organisms (including humans) by precisely manipulating genes.89 This could lead to novel disease treatments and new types of genetical y engineered products (such as genetical y engineered biofuels), while enabling the nascent field of synthetic biology—designing DNA from scratch to produce desired traits.90 The potential economic impact of next-generation gene sequencing in the applications that we have sized in health care, agriculture, and the production of substances such as biofuels could be $700 bil ion to $1.6 tril ion a year by 2025. About 80 percent of this potential value would be realized through extending and enhancing lives through faster disease detection, more precise diagnoses, new drugs, and more tailored disease treatments (customized both to the patient and to the disease). In agriculture, analyzing plant genomes could lead to more advanced genetical y modified (GM) crops and further optimize the process of farming by tailoring growing conditions and farming processes to a seed’s genetic characteristics. Furthermore, it may be possible to create high-value substances such as biofuels by modifying simple organisms such as E. coli bacteria. Easy access to gene-sequencing machines could not only put powerful genetic technology in the hands of researchers and physicians, but could also create a global community of co-creators that might advance biotechnology in unforeseeable ways, as hobbyists propel ed the microcomputer revolution. The technical chal enges inherent in next-wave genetic engineering technology are great but may be less formidable than the social, ethical, and regulatory 89 The fundamental chal enge of genomics and, arguably, biology itself, involves clearly understanding the processes that link DNA, RNA, proteins, and metabolics to determine cel ular and organism behavior, a field of discovery often referred to as “omics.” 90 DNA can be modified through recombinant technology, combining DNA from two or more existing genomes, or through a process of DNA synthesis that involves chemical y printing DNA (often cal ed synthetic biology). Synthetic biology is at a relatively early stage of technological development; furthermore, precise editing of DNA is not required to achieve much of the impact made possible by next-generation genomics. For these reasons, we do not include synthetic biology in our estimated potential economic impact in this report.
88 concerns it may generate. While this technology has the potential to create huge benefits for society, it comes with an equal y impressive set of risks. Genetical y modified organisms could interfere with natural ecosystems, with potential y disastrous results including loss of species and habitats. Genomic technology raises privacy and security concerns related to the potential theft or misuse of personal genetic information stored on computers. And while the potential for widespread access to sequencing and, eventually, DNA synthesis technology wil create opportunities for innovators, it also raises the specter of bioterrorism. Moreover, this technology could wel unfold in a regulatory vacuum: governments have yet to address major questions concerning who should own genetic information, what it can be used for, and who should have access to next- generation genomic capabilities. DEFINITION Next-generation genomics can be described as the combination of next- generation sequencing technologies, big data analytics, and technologies with the ability to modify organisms, which include both recombinant techniques and DNA synthesis (that is, synthetic biology). Next-generation sequencing represents newer, lower-cost methods for sequencing—or decoding—DNA. It encompasses the second- and third-generation sequencing systems now coming into widespread use, both of which can sequence many different parts of a genome in paral el. The rate of improvement in gene-sequencing technology over the past decade has been astonishing. When the first human genome was sequenced in 2003, it cost nearly $3 bil ion and took 13 years of work by teams of scientists from all over the world col aborating on the Human Genome Project. Now a $1,000 sequencing machine could soon be available that wil be able to sequence a human genome in a few hours. In fact, over the past decade, the rate of improvement in sequencing speed has exceeded Moore’s law, the famously fast rate of performance improvement achieved by computer processors. This improvement in performance has been achieved by creating highly paral el systems that can sequence mil ions of DNA base pairs in a very short time. As the DNA is read, the process generates massive data, which are passed on to powerful computers for decoding. Thus, progress in genomics and computing speed are evolving in tandem—a development that has been referred to as “wet” science meeting “dry” science. This advance in DNA sequencing speed (along with simultaneous reductions in cost) promises to accelerate the process of biological discovery. Historical y, biological research has relied largely on hypothesis-driven, trial-and-error testing. This approach is very time consuming and difficult, so scientists’ understanding of which genes drive specific outcomes (such as diseases) remains very limited. With growing access to large samples of ful y sequenced genomes, researchers can employ more broad-based methods, performing correlation analysis on big data sets of sequenced genomes together with patient data, and testing combinations of genes, diseases, and organism characteristics to determine which genes drive which outcomes. These big data experiments can include data on genealogy, clinical studies, and any other statistics that could help link genotype (DNA) to phenotype (organism characteristics or behavior). Armed with this information, it could be possible to better identify and diagnose people at high risk for conditions such as heart disease or diabetes, al owing earlier, more effective intervention.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 89 Next-generation sequencing also makes personalized medicine possible. Individual patients possess unique genomes and can be affected differently by the same disease or therapy. The ability to genetical y sequence al patients, along with the viruses, bacteria, and cancers that affect them, can al ow for better matching of therapy to the patient. Sequencing can also help physicians understand whether a set of symptoms currently treated as a single disease is, in fact, caused by multiple factors. Advanced genomics wil also facilitate advances in agriculture. Farmers might be better able to optimize soil types, watering schedules, crop rotations, and other growing conditions based on a more complete understanding of crop genomes. It may also be possible to produce genetical y modified crops that can grow in locations where soil conditions and access to water cannot be easily improved; crops that can thrive in colder, drier climates; or crops that generate a larger portion of their weight as food. Crops might also be modified to serve as better raw materials for the production of biofuels. Modified animals in our food supply might also be on the horizon. For example, one US company is seeking approval for an Atlantic salmon modified with an eel gene that enables it to reach maturity in half the normal time.91 Final y, next-generation genomic technology could be used to modify the DNA of common organisms to produce valuable substances. Using synthetic biology or even standard, wel -established recombinant techniques, the metabolic systems of certain organisms can be modified to produce specific substances, potential y including fuels, pharmaceuticals, and chemicals for cosmetics. POTENTIAL FOR ACCELERATION Next-generation genomics has the potential to give humans far greater power over biology, al owing us to cure diseases or customize organisms to help meet the world’s need for food, fuel, and medicine. With world population heading toward eight bil ion in 2025, there is a growing need for more efficient ways to provide fuel for heat, electricity generation, and transportation; to feed people; and to cure their ailments. Meanwhile, populations are aging in advanced economies. By 2025 approximately 15 percent of the world’s population wil be 60 years of age or older, multiplying health-care chal enges.92 Next-generation genomics can address these needs. The falling cost of genome- sequencing technology wil accelerate both knowledge and applications. In genomics, the relevant unit of performance measurement is the time and cost per sequenced base pair (the basic units of DNA). With newer generations of sequencing technology, the cost of sequencing a ful human genome has fal en to around $5,000, but the $1,000 genome is widely expected to be achieved within the next few years. Counsyl, a Silicon Val ey company, already offers a $600 genetic test that can screen children for more than 400 mutations and 100 genetic disorders.93 91 Andrew Pol ack, “Engineered fish moves a step closer to approval,” The New York Times, December 21, 2012. 92 David Hutton, Older people in emergencies: Considerations for action and policy development, World Health Organization, 2008. 93 Kim-Mai Cutler, “Through dirt-cheap genetic testing, Counsyl is pioneering a new bioinformatics wave,” TechCrunch.com, April 23, 2013.
90 Cancer is a genetic disease that is caused when mutated cel s grow out of control. Sequencing is already being used to tailor treatments that are customized to the genome of the patient and the mutated genome of the tumor. Studies have shown how specific cancer-causing mutations correlate with responses to different cancer treatments and there is a healthy pipeline of bio treatment and diagnostic drugs. However, given the rate at which drugs fail during testing, it is not likely that the number of drugs used with companion diagnostics over the next five years wil rise rapidly.94 Oncology remains at the forefront of genetic research and development in medicine, but applications for other types of diseases are on the radar. Researchers are also focusing on mutation-based links to widespread diseases such as cardiovascular disease to identify how different genomes correspond to different responses to therapies. GM crops are playing an increasingly important role in improving agriculture in developing economies. The total area planted in GM crops has risen from 1.7 mil ion hectares in 1996 to more than 170 mil ion hectares in 2012, and for the first time farmers in developing economies planted more hectares of GM crops than did farmers in advanced countries. Planting of GM crops grew 11 percent in developing economies in 2012, more than three times the rate of such planting in advanced economies. Next-generation genomics could enable the creation of even more advanced varieties with even greater potential value. Synthetic biology is stil in a very early stage of development, but could become a source of growth. If the process can be perfected, modifying organisms could become as simple as writing computer code. While the technology is new, there is already evidence for applications in science and business. For example, a research team at Ginko Bioworks in Boston is working on developing the biological equivalent of a high-level programming language with the goal of enabling large-scale production of synthetical y engineered organisms. Companies are beginning to invest in synthetic biology capabilities. Joule Unlimited and Algenol Biofuels, for example, have created demonstration plants that can produce high-value substances using synthetical y engineered organisms—diesel fuel in the case of Joule Unlimited and ethanol at Algenol. However, synthetic biology remains chal enging, with high up-front capital investments required and difficulties in economical y scaling production. ECONOMIC IMPACT In the applications we assessed, we estimate that next-generation genomics have a potential economic impact of $700 bil ion to $1.6 tril ion per year by 2025. We estimate the impact of disease prevention and treatment applications that we size could be $500 bil ion to $1.2 tril ion per year in 2025, based on extended life expectancy stemming from better and faster disease diagnosis and more tailored treatments (Exhibit 9). In particular, new technology has the potential to improve treatment of genetical y linked diseases such as cancer and cardiovascular diseases, which currently kil around 26 mil ion patients per year. 94 Personalized medicine: The path forward, McKinsey & Company, March 2013.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 91
Exhibit 9 Sized applications of next-generation genomics could have direct economic impact of $700 billion to $1.6 trillion per year in 2025 Potential economic impact of sized Sized applications in 2025 Estimated scope Estimated potential Potential productivity applications $ tril ion, annual y in 2025 reach in 2025 or value gains in 2025 Estimated deaths Patients with access to Extended life expectancy from relevant relevant treatment – Cancer: 0.5–2 years2 diseases – Cancer: 20–40% – Cardiovascular: Disease 0.5– – Cancer: 12 mil ion – Cardiovascular: 1 year treatment 1.2 – Cardiovascular: 15–40% – Type 2 diabetes: 23 mil ion – Type 2 diabetes: 1 year3
– Type 2 diabetes: 20–40% Value of prenatal
4 mil ion Access to prenatal screening4
160 mil ion newborns genetic screening: – Developed world: Substance 0.1– – Developed world: 100% $1,000 production 0.2 – Less-developed: – Less-developed: $200 30–50%1 60 bil ion gallons per Ethanol: 20–40% of world 15–20% cost saving in year of ethanol production ethanol production 0.1– 350–500 bil ion Diesel: 2–3% of world 150–200% price Agriculture 0.2 gallons per year of production premium for diesel diesel 30–70% CO2 reduction from fuels over life cycle Other $1.2–1.3 tril ion worth 60–80% of agricultural 5–10% increase in yields potential of major crops (wheat, production improved due to process applications maize, rice, soybeans, using genomics data optimization (not sized) barley, tomatoes) 20–80% of current 5–10% increase in yields genetical y engineered from use of advanced crops to be further genetical y engineered Sum of sized enhanced crops potential 0.7– economic 1.6 impacts 1 Developing economies excluding the least developed. 2 Wil vary across cancer types. 3 We take into account the overlap of diabetes- and cardiovascular-related deaths. 4 Measured by parents’ wil ingness to pay. NOTE: Estimates of potential economic impact are for some applications only and are not comprehensive estimates of total potential impact. Estimates include consumer surplus and cannot be related to potential company revenue, market size, or GDP impact. We do not size possible surplus shifts among companies and industries, or between companies and consumers. These estimates are not risk- or probability-adjusted. Numbers may not sum due to rounding. SOURCE: McKinsey Global Institute analysis Some 14 mil ion new cases of life-threatening cancers can be expected to be diagnosed worldwide in 2025. Determining how many of these patients could have longer lives or better quality of life due to more effective treatment based on next-generation genomics is not straightforward. Most cancers stil may not be curable even after sequencing identifies the genetic mutations that trigger disease. For some cancers, however, the process of identifying the mutations involved and then developing targeted therapy is moving ahead. For example, Herceptin, a breast cancer drug, acts only on tumors that contain cancer cel s that, because of a gene mutation, make more of the HER2 tumor-creating protein than normal cel s. Studies have shown that Herceptin can decrease fatalities by, for example, reducing the risk of recurring tumors. Some industry leaders believe that eventual y most types of cancer could be treated with targeted therapies based on next-generation genetic sequencing.95 To estimate the potential economic impact of these improved diagnostic and tailored treatment methods, we estimate the value to the patient of the extended life that might result. Based on the as 9 sessm ent of cancer experts, we estimate that genomic-based diagnoses and treatments can extend lives of cancer 95 Personalized medicine, McKinsey & Company, March 2013.
92 patients by six months to two years in 2025. We further estimate that 20 to 40 percent of patients would have access to such care in 2025 . We note that any estimates of success rates are highly speculative, given the state of development of these therapies. Longer term, advanced genomics may offer tremendous potential to develop personalized treatments for cardiovascular disease. Every patient responds differently to the mix of medicine they are exposed to, and today high-risk patients are often treated with preventive medications with dosages adjusted on a trial- and-error basis, creating high risks. While the technology is stil in early stages, genetic testing could help doctors determine dosages and mixes of substances more precisely. Also, screening can enable customized preventive routines (lifestyle changes, for example), as wel as tailored treatments. Based on expected growth rates in cardiovascular disease, 23 mil ion people could be expected to die of cardiovascular disease in 2025.96 For the purpose of sizing potential impact, we assume that 15 to 40 percent of patients could receive and benefit from genetic-based care and, on average, have one year of extended life. Another major target for genetic medicine is type 2 diabetes, a growing health problem, especial y in advanced economies. Based on current rates of diabetes incidence, some three mil ion deaths could be caused by type 2 diabetes and related complications in 2025.97 Genome sequencing could enable the creation of treatments that could control the disease more effectively and better reduce the risk of death than the currently imprecise science of daily insulin use. We estimate that 20 to 40 percent of patients could have access to such treatment and might have increased life expectancy of one year.98 Other areas in which genetic sequencing holds promise, but for which we have not built estimates, include immunology and transplant medicine, central nervous system disorders, pediatric medicine, prenatal care, and infectious diseases. Next-generation gene sequencing has application in prenatal care. Sequencing a fetus’s DNA would make it possible to predict the health of the baby more accurately than current tests can. Surveys show that parents in advanced economies would be wil ing to pay $1,000 to have their baby’s genome sequenced.99 Assuming close to 100 percent adoption in advanced economies and 30 to 50 percent adoption rates in less-developed economies (excluding least-developed nations), this testing could generate value of approximately $30 bil ion per year in 2025. Another source of potential economic impact could arise from altering the metabolism of common organisms such as E. coli and yeast to create biofuels; this could be less expensive and require less energy and other inputs than creating biofuels from plants. Production costs could be as much as 15 to 20 percent lower for ethanol produced in this manner. It is possible that the cost of producing diesel using this technology could reach parity with traditional diesel by 2025, and since biodiesel commands a price premium even today (due 96 Cancer Fact Sheet number 297, World Health Organization, January 2013. 97 Diabetes, Fact Sheet number 312, World Health Organization, March 2013. 98 There is an overlap in potential economic impact from diabetes- and cardiovascular- related deaths. 99 A. B. Caughey et. al., “Assessment of demand for prenatal diagnostic testing using wil ingness to pay,” Obstetrics and Gynecology, volume 103, number 3, March 2004.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 93 to factors like government policy and environmental concerns), this new source of biodiesel could see significant demand. We estimate that the price premium over traditional diesel might be as much as 150 to 200 percent, and that micro- organism produced biodiesel could potential y replace 2 to 3 percent of traditional diesel consumption. Producers claim that these fuels derived from microbes could contribute 70 to 90 percent less in CO emissions in their production 2 and use than traditional fuels. We estimate that the potential impact could be $100 bil ion to $200 tril ion in 2025, including the savings associated with lower production costs, the value of lower CO emissions, and the value of the fuel itself. 2 In agriculture, next-generation genomics has the potential to both raise productivity in places where food is in short supply and conserve water. Advances in genetic modification of seeds could increase yields by making crops more drought- and pest-resistant. Genomics can also provide information that can be used to optimize crops for specific soils and climates and guide precision farming practices such as “fertigation” (a process in which the exactly necessary amounts of water and fertilizer are delivered to crops). We assume potential increases of 5 to 10 percent from optimized processes and 5 to 10 percent from new genetical y modified seeds, leading to a potential economic impact of $100 bil ion to $200 bil ion per year in 2025. BARRIERS AND ENABLERS There is stil much that scientists do not understand about genomics. Deciphering the interrelationships between genes, cel ular mechanisms, organism traits, and environment is a complex undertaking that next-generation gene sequencing can speed up. But that work is only just beginning. Fast and cheap sequencing is stil a new technology, and much of the initial genome sequencing will likely have unknown or little immediate impact. Understanding and applications wil grow once many genomes have been sequenced and sample sizes are big enough to enable advanced analytics. While next-generation genomics technology could speed up this learning process dramatical y, it is less clear how quickly this will lead to breakthroughs in understanding biology. What is more likely to slow progress, however, are the many unresolved regulatory and ethical issues that this technology poses. One issue is the ownership of the data of sequenced genomes, which wil be a very valuable resource for performing analyses and testing pharmaceutical treatments, but which might not be available if patients own the data regarding their own genomes and are not wil ing to share it. In the fal of 2013, the US Supreme Court is scheduled to hand down a key ruling on whether pharmaceutical companies can patent human cel s. There are also concerns regarding the confidentiality of patient DNA information: can it be used by health insurers to deny coverage or raise rates, and should patients be given al the information about disease-linked mutations found in their genomes that might someday lead to il ness? If these questions are slow in being addressed or not addressed adequately, progress could be delayed, potential y by public resistance. There is also widespread public apprehension about the possible unintended consequences of altering plant and animal DNA. The European Union’s 1998 ban on genetical y modified corn remains in effect, and many consumers are concerned about the possible effects of “Frankenfood” on environments, biodiversity, and human health. Advanced recombinant technology and synthetic biology could certainly heighten such concerns. Regulators have imposed limits
94 on research on modified organisms, restricting them to closed environments. The continuation or strengthening of these types of restrictions could limit the potential economic impact of advanced genomics. IMPLICATIONS By 2025, continued advancement in gene sequencing speed and cost, along with equal y rapid advances in the ability to understand and manipulate biological information, could create tremendous opportunities and risks for technology providers, physicians, health-care payers, biotechnology and pharmaceutical companies, entrepreneurs, and societies. It is possible that genetic sequencing could become standard practice during medical exams by 2025. If it does, it could create major opportunities for companies and startups to manufacture and sel gene-sequencing equipment, along with the various supporting systems and tools that could be required (including big data analytics tools). The early entrants into this market could have the opportunity to define major industry standards and norms, including sequencing approaches, data standards, and integration with electronic health records. They wil also need to win over payers, such as insurance companies and governments, by clearly demonstrating cost-effective efficacy improvements; the technology wil also need to become user-friendly for physicians. Insurers and other health-care payers (for example, state health insurance systems) wil have a large interest in shaping how next-generation genomics and the resulting data are used. Improved treatments, reduced side effects, and reduced waste (due to avoiding incorrect diagnoses and treatments) could help reduce payer costs, which could provide the incentive for payers to subsidize routing genome screening. In some countries payers wil also need to convince regulators and patients that genomic data wil not be used against individual patients. For biotech and pharmaceutical companies, the ability to sequence more material more quickly and to use the growing body of genetic data to isolate (or engineer) the best candidate substances for drug development has the potential to raise productivity and lower costs for new drugs and therapies. This could significantly impact the economics of drug discovery and testing. Simultaneously, the barriers to entry into biopharmaceuticals could fal when research becomes less capital intensive and cycles for development decrease. Next-generation genomics could drive a major wave of entrepreneurship. Alongside companies that are looking to produce and sel gene-sequencing systems, there are already a growing number of startups and laboratories offering home DNA testing, including results that identify predispositions for known genetic diseases and information on ancestry. Fast, cheap sequencing is making these types of services possible; however, as these technologies become widely available, specialized testing services may have a limited market. Other entrepreneurs are already coming to market with new, unexpected solutions based on next-generation sequencing. For example, a synthetic biology startup that used the peer-to-peer funding site Kickstarter to raise capital has produced a synthetical y engineered light-emitting plant, which it says could lead to a new source of lighting.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 95 Policy makers wil have many issues to address surrounding the applications of genetic science. Governments have supported the development of genomics through investments in research, but have not been as forward-thinking when it comes to crafting policy. Governments can play a critical role in helping next- generation genomics live up to its potential to save lives, feed people, and provide fuels that wil be less harmful to the environment. One possible step could be supporting independent research investigating questions regarding the environmental and health risks and benefits of genomic applications. Governments can also work on regulations and initiatives to enable the success of genome-based advances. For example, the regulatory environment for drugs and diagnostics is likely to have a significant impact on the evolution of personalized medicine. Most experts believe that regulation has not kept pace with the rapid advances in the field of personalized medicine. A more far-sighted regulatory approach could balance many of the objections, including concerns regarding personal privacy, with the potential benefits of these technologies. This would give next-generation genomics researchers the opportunity to continue developing these technologies. In addition to clarifying rules about ownership of DNA data and confidentiality, governments can facilitate the accumulation of genetic information. Since 2006, the US National Cancer Institute and the National Human Genome Research Institute have been compiling the Cancer Genome Atlas, a project with the goal of col ecting al data about what mutations have been linked to cancers. It may soon be possible for governments to sponsor a central database of mil ions of genome sequences and make the information accessible to researchers (with proper precautions). Perhaps the thorniest concern in the near future is prenatal genome sequencing. Prenatal genetic screening raises the specter of eugenics: wil parents end pregnancies for reasons other than serious deformities and other congenital medical conditions? Ever since the creation of Dol y the sheep proved that cloning is possible in 1996, genetic engineering has inspired both visions of a better world and concerns about the risks of such advances. Recently, scientists have revealed that they have successful y inserted mitochondrial DNA into the egg cel s of women who have had trouble conceiving. The procedure has been used in 30 successful pregnancies, producing babies with genes from the child’s two biological parents and the mitochondrial DNA donor; in effect, these children have three biological parents.100 This particular modification was performed to aid in conception, but it could also represent the first step on the path to manipulating human DNA to produce babies with “desirable” traits. 100 Michael Hanlon, “World’s first GM babies born,” Daily Mail, April 25, 2013.
96 8. Energy storage Since the late 1700s—a century before electricity became widely used—scientists have been working on ways to store electrical energy. Today, the technology of energy storage is advancing rapidly and being applied in new ways, creating significant potential for impact and disruption. Improved lithium-ion batteries are already powering electric and hybrid vehicles, as wel as bil ions of portable consumer electronics products, including mobile Internet devices. Over the coming decade, battery-powered vehicles could become cost competitive with vehicles powered by internal-combustion engines, saving energy and reducing CO emissions. Energy storage could also help bring electricity to remote areas in 2 developing countries and boost the efficiency and quality of the electric grid while helping to reduce CO emissions. 2 The potential economic impact of improved energy storage could be $90 bil ion to $635 bil ion per year by 2025. More than half of this impact could be driven by electric and hybrid vehicle adoption. Distributed energy (that is, using batteries to bring power to areas where wiring or reliable supply is not available) may have a relatively smal direct economic impact but have a transformative effect on the lives of more than one bil ion people who currently live without electricity. Grid applications—placing storage capacity on the grid to reduce the cost of meeting peak demand and to facilitate feeds from wind or solar generators—may also have relatively modest impact by 2025, barring major technical breakthroughs in battery cost and performance that would improve gains. In this chapter, we analyze the potential acceleration of energy storage use and its potential impact assuming the scenario that has the most realistic potential by 2025. However, it is possible to envision a scenario in which distributed storage has a faster rate of adoption and greater economic impact as a result of fal ing battery costs and higher energy prices, which could lead to significantly increased adoption of renewable power sources, particularly solar (see Chapter 12, “Renewable energy”) and drive demand for grid storage. While this scenario is more likely to occur after 2025, it nonetheless deserves mention. DEFINITION Energy storage systems convert electricity into a form that can be stored and converted back into electrical energy for later use, providing energy on demand. This enables utilities, for example, to generate extra electricity during times of low demand and use it to augment capacity at times of high demand. Today, about 3 to 4 percent of the electricity that is produced by utilities worldwide is stored, almost al of it through a technique cal ed pumped hydro-electric storage (PHES), which involves pumping water uphil during times of low demand or low cost and releasing it downhil to turn power-generating turbines during times of high
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 97 demand and high cost. PHES currently accounts for approximately 120 gigawatts of storage capacity.101 Batteries in their various forms constitute the most widely known energy storage technology and the main focus of our analysis. Lithium ion (Li-ion) batteries are widely used in consumer electronic devices such as laptop PCs, as wel as in electric and plug-in hybrid vehicles. The Li-ion battery market is expected to double in the next four years to $24 bil ion in global revenue.102 Significant performance and cost improvements are also expected in Li-ion batteries over the coming decade. Prices for complete automotive Li-ion battery packs could fal from $500 to $600 per kilowatt-hour today to about $160 per kilowatt-hour in 2025, while cycle life could increase significantly at the same time, potential y making plug-in hybrids and electric vehicles cost competitive with traditional internal combustion engine vehicles on a total cost of ownership basis. The average cost of owning and operating Li-ion batteries for utility grid applications (a function of multiple variables including battery prices and cycle life) could fal from $500 per MWh to between $85 and $125 per MWh by 2025. This could make Li-ion batteries cost competitive for some grid applications and for providing distributed energy, based on the levelized cost of electricity (LCOE), a standard measure of electricity costs.103 Other important energy storage technologies include molten salt, flow cel s, fly wheels, supercapacitors, and even conventional lead acid batteries (including recycled batteries). Other promising battery technologies to watch that are currently under development but may not be commercial y viable by 2025 include liquid metal, lithium-air, lithium-sulfur, sodium-ion, nano-based supercapacitors (see Chapter 10, “Advanced materials”), and energy cache technology. Compressed air energy storage (CAES) is a fairly mature energy storage technology useful for utility grid applications. Similar to PHES, CAES typical y uses natural formations, but instead of pumping water uphil , air is pumped into caverns, salt domes, or other underground spaces and maintained under pressure until it is released to help drive a turbine. By 2025, the next generation of PHES and CAES energy storage could enable construction methods that are less dependent on natural y occurring geographic formations (making use instead of sea water, mine shafts, and cargo containers, for example), as wel as variable- speed turbines, giving more output control and higher round-trip efficiencies. POTENTIAL FOR ACCELERATION Energy storage systems play an important role in integrating alternatives to fossil fuels into the energy mix and also can help improve the reliability of the electric supply and bring electricity to new users. With growing energy demand and growing concerns over CO emissions and climate change, there is growing 2 demand for less harmful means of energy production. Today, 13 bil ion tons 101 Electricity storage, International Energy Agency/Energy Technology Systems Analysis Program and International Renewable Energy Agency technology policy brief E-18, April 2012. 102 Malavika Tohani, Global lithium-ion battery market: Growth trends and application analysis, Frost & Sul ivan, February 2013. 103 Cost is often measured as the levelized cost of electricity (LCOE), the constant unit cost (per KWh or MWh) of electricity generated by different sources using a present value payment stream of the total cost of capital, return on investment, operating costs, fuel, and maintenance over a technology’s useful life. This measure is useful for comparing the prices of technologies with different operating characteristics.
98 of CO are released annual y from electricity generation. Seven bil ion tons are 2 released annual y through transportation.104 The energy and transportation sectors are beginning to add more sustainable energy sources and, in both sectors, these efforts rely on energy storage: on the energy grid, storage systems can help accommodate electricity from renewable sources such as solar and advanced batteries make electric and partial y-electric vehicles possible. Several other factors are leading to increased interest in energy storage. Energy storage costs have declined in recent years and are expected to decline even more by 2025—particularly for Li-ion batteries—although experts disagree as to exactly how much. This has enabled increased adoption of hybrid and battery- operated vehicles, as wel as higher-performance portable consumer electronics. Improved energy storage technologies could help rapidly growing developing economies meet their energy needs. China’s electricity consumption is growing by 11 percent per year, India’s by 5 percent, and Africa’s by 4 percent.105 Advanced energy storage technologies may be able to bring power to areas that are not currently wired and may not be for many years to come, and can be used within power grids to stretch capacity until new infrastructure is built. Utilities in advanced economies may also invest more in energy storage during the coming decade in order to deal with peak capacity issues, to accommodate renewables (taking electricity from solar and wind farms, for example), and as part of smart grid instal ations. The rising use of electricity from solar and wind, which are intermittent sources of power, wil likely require new forms of energy storage.106 To meet peak demand without adding permanent new capacity, utilities have a range of choices, including PHES, CAES, and other non- battery technologies. Battery storage could become more competitive in these applications, but may apply only in limited circumstances. Battery storage in smart grid applications can help with frequency regulation and guaranteed peak power services. Major advances in battery technology are occurring in important components, which could double battery capacity over the next 10 to 15 years.107 Batteries have three elements: a positive terminal (the cathode), a negative terminal (the anode), and an electrolyte (a chemical medium that al ows the flow of electrical charge between the cathode and anode). Next-generation cathodes incorporate “layered-layered” structures, eliminating dead zones that impede ion transfer. Silicon anodes could increase cel capacity by 30 percent compared with graphite anodes (although they are currently susceptible to cracks).108 In addition, researchers are trying to identify cathode-electrolyte pairs that can sustain higher voltages, thereby boosting capacity. These advances, combined with increased production efficiency, are widely expected to significantly reduce the LCOE of batteries over the coming decade. 104 World energy outlook 2011, International Energy Agency, November 2011. 105 Annual energy outlook 2012, US Energy Information Administration, April 5–May 2, 2013. 106 Many renewable energy sources, such as solar and wind power, provide intermittent power (when wind velocity drops or the sun is obstructed by clouds, for example). 107 Russel Hensley, John Newman, and Matt Rogers, “Battery technology charges ahead,” The McKinsey Quarterly, July 2012. 108 Ibid.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 99 POTENTIAL ECONOMIC IMPACT The economic impact of energy storage technologies in the applications we analyzed has the potential to reach $90 bil ion to $635 bil ion annual y in 2025 (Exhibit 10). This value could arise from three primary applications: electric and hybrid vehicles, distributed energy, and utility grid storage. While energy storage for consumer electronics also has significant economic value, a major portion of this value is effectively included within our estimated potential economic impact for mobile Internet technology (see Chapter 1), and therefore we do not include it here.
Exhibit 10 Sized applications of energy storage could have economic impact of $90 billion to $635 billion per year in 2025, including consumer surplus Potential economic impact of sized Potential Sized applications in 2025 Estimated scope Estimated potential productivity or applications $ bil ion, annually in 2025 reach in 2025 value gains in 2025 115 mil ion passenger 40–100% of Fuel price: $2.80– vehicles sold vehicles sold in 7.60 per gallon
Over 1 bil ion vehicles 2025 could be 0.22 KWh per mile
in the market electric or hybrid fuel efficiency for EVs
13,000 TWh electricity 35–55% adoption $0.75–2.10 per Electric consumption in with solar and KWh value of and 20– emerging markets battery combination uninterrupted hybrid 415 2–70 hours per month 35–55% of power supply to an vehicles without electricity companies in enterprise Africa, Middle East, $0.20–0.60 per and South Asia own KWh value per Stabilizing diesel generators household 25– electricity 100 access 60–65% rural 50–55% adoption $0.20–0.60 per Distri- electrification rate based on number of KWh value per buted 1.2 bil ion people people projected to household for energy without electricity earn above $2 per direct lighting, TV, Electrifying 0– access day and radio benefits new areas 50 60 KWh monthly electricity requirement of average household 27,000–31,000 TWh 100% technology $30 per MWh Frequency 25– global electricity adoption, more weighted average regulation 35 consumption efficient, and cost frequency- 1.5% electricity competitive with regulation price production reserved incumbent solutions for frequency regulation Utility Peak load 10– 2.5% additional grid shifting 25 reserved for renewable integration 12% of total electricity 10–20% adoption of $65–80 per MWh Infra- production possible to energy storage, between non- structure ~10 shift given costs renewable peak deferral 850 mil ion tons compared with and base load additional CO2 release combined cycle gas $45–65 per MWh turbines between peak and average wind price Other $30–45 per MWh potential between peak and applications average solar price (not sized) $295 bil ion per year 15% adoption Possible deferral of Sum of sized investment in based on share of infrastructure potential 90– infrastructure T&D transmission lines investment by economic 635 deferral economical for 2.5 years impacts 10% spent to reduce energy storage congestion NOTE: Estimates of potential economic impact are for some applications only and are not comprehensive estimates of total potential impact. Estimates include consumer surplus and cannot be related to potential company revenue, market size, or GDP impact. We do not size possible surplus shifts among companies and industries, or between companies and consumers. These estimates are not risk- or probability-adjusted. Numbers may not sum due to rounding. SOURCE: McKinsey Global Institute analysis 10
100 Electric and hybrid vehicles Based on an expected doubling of lithium-ion battery capacity in the next decade, electric-powered vehicles (EVs) could become cost competitive with internal combustion engine (ICE) vehicles by 2025 on a total cost of ownership basis. The cost of energy from lithium-ion battery packs (which usual y include battery cel s, a battery management system to control cel balancing during cycles, and thermal protection) could drop to $165 per KWh by 2025, from $560 per KWh in 2011.109 This decrease could be the result of improvements in battery capacity and manufacturing efficiency, as wel as fal ing component costs. Cell capacity could improve by 110 percent due to innovations such as layered-layered cathodes and silicon anodes and as volumes rise, we estimate that margins for suppliers of raw materials and parts could fal by 20 to 40 percent from today’s levels, reducing costs for automakers. Given these factors, the potential economic impact of energy storage for electric vehicles could be approximately $20 bil ion to $415 bil ion annual y by 2025. For the purposes of calculating the magnitude of this impact, we estimated the number of battery-powered vehicles that could be sold annual y in 2025 based on historic adoption trends and expected battery improvements that would alter the cost comparison between EVs and ICEs. For this analysis, we assume that the adoption of hybrid, plug-in hybrid, or battery-powered vehicles wil be highly dependent on retail fuel prices. Using these assumptions, and accounting for wide regional fuel price variation (plus or minus 50 percent of current fuel prices), 20 to 40 percent of new cars bought global y in 2025 could be hybrid electric vehicles. This could occur if the total cost of ownership of hybrid vehicles makes them cost competitive with ICEs when fuel prices are $2.85 per gal on or more. This also assumes that hybrids use electricity about 40 percent of the time. As much as 55 percent of new cars could be plug-in hybrid vehicles by 2025, which could be cost competitive with internal-combustion and hybrid vehicles at fuel prices above $5.50 per gal on (these cars use electricity during approximately 75 percent of driving time). Final y, up to 25 percent of new cars could be al - electric, which would be competitive at fuel prices above $9.50 per gal on.110 Electricity distribution in underserved markets Bringing electricity to developing economies is a significant chal enge. In many nations, power plants operate below their rated capacities due to fuel shortages or dropping water levels in hydro-electric systems. In addition, there may be limited or highly unreliable grid infrastructure, particularly transmission and distribution lines reaching rural areas. For distributed energy, the total potential economic value of energy storage could be between $25 bil ion and $150 bil ion in 2025. This estimate is based on two sources of impact: improving the reliability of electric power in developing economies and bringing electric power (at least on a part-time basis) to the poorest citizens living in the most remote areas. 109 Russel Hensley, John Newman, and Matt Rogers, “Battery technology,” The McKinsey Quarterly, July 2012. 110 These are ranges of maximum market share per electric-vehicle type and are interdependent (i.e., if plug-in hybrids achieve 55 percent share in 2025, plug-in hybrids cannot have 40 percent share). Therefore, these shares do not add up to more than 100 percent in any scenario.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 101 Unstable electricity supply About 43 percent of the electricity consumed in developing economies is used by industry.111 In a survey undertaken by the World Bank, 55 percent of Middle Eastern and North African firms, 54 percent of South Asian firms, and 49 percent of sub-Saharan African firms identified electricity as a major constraint to doing business. The smal est firms, which cannot afford backup generators, are most affected by erratic electricity supply; however, large firms stil feel the impact, since they are forced to commit resources to generating capacity in order to keep running.112 The potential for improvement of electricity reliability in developing economies is vast: developing economies consume 13,000 TWh of electricity annual y, even though outages in such nations are persistent, lasting from two to 70 hours per month on average.113 Using energy storage to create distributed sources of additional power could have an economic impact of $25 bil ion to $100 bil ion annual y in 2025 (less the cost of storage) by preventing outages. We base this estimate on an adoption rate of 35 to 55 percent. We place the value of uninterrupted electricity at 75 cents to $2.10 per kilowatt-hour for businesses and of 20 to 60 cents for consumers, based on developing country outage reports.114 New electricity supply Today only 63 percent of rural populations in emerging markets have access to electricity. More than one bil ion people could stil be without electricity in 2025, based on population growth and current levels of electrification.115 In remote, sparsely populated areas, local energy generation is often the only solution for electricity access. The value of providing electricity access to remote areas in developing economies could be as much as $50 bil ion annual y in 2025. We base our estimate on a solar-plus energy storage solution, providing 60 KWh of electricity per month per household for lighting and some television, cell phone charging, radio, fan, and iron usage. The value to rural households of this electricity access is estimated to be between 20 to 60 cents per KWh (based on World Bank and IEG projections, and an estimated 55 percent adoption rate, based on the proportion of rural populations that could potential y afford system lease fees).116 Grid storage Today, electricity is generated seconds before it is used on the grid. As a result, the electricity industry must invest in and maintain capacity for peak usage— usual y during the heat of summer as mil ions of air conditioners run at full power—even though peak demand is infrequent. Energy storage could enable peak load shifting (tapping additional sources of supply during times of peak demand), higher utilization of existing grid infrastructure, and efficient balancing of smal fluctuations in power output, as wel as providing temporary power in the 111 Russel Hensley, John Newman, and Matt Rogers, “Battery technology,” The McKinsey Quarterly, July 2012. 112 World Bank Group, Enterprise surveys: What businesses experience. 113 World energy outlook 2011, International Energy Agency, November 2011. 114 Musiliu O. Oseni, Power outages and the costs of unsupplied electricity: Evidence from backup generation among firms in Africa, University of Cambridge research paper. 115 World energy outlook 2011, International Energy Agency, November 2011. 116 The welfare impact of rural electrification: A reassessment of the costs and benefits, Independent Evaluation Group (IEG) and World Bank, 2008.
102 event of an outage. However, for these benefits to be realized, energy storage must be cost competitive with other methods of addressing these issues, such as combined cycle gas turbine technology (the type of generating plant used for peaker applications) and demand-side management (that is, getting consumers and businesses to voluntarily reduce use during peak periods). Even by 2025, battery storage may not be competitive in many circumstances. Given these limitations, grid-based energy storage could have a moderate economic impact of $45 bil ion to $70 bil ion annual y in 2025. This value could be realized primarily from three applications: frequency regulation, peak load shifting, and deferral of investments in new transmission and distribution infrastructure. Frequency regulation A wide range of devices, such as motors used in manufacturing, rely on constant frequency of alternating current electricity (50 to 60 Hz for most of the world). When generation and load are out of balance, frequency deviates from its set point. Significant load increases cause frequency to slow and voltage to drop. Similarly, frequency increases are caused by loss of load, which can happen with renewables (when a cloud passes over the sun during solar production, for example). Conventional generating facilities powered by gas or coal provide their own frequency regulation —a constant flow—by setting aside a portion of generating capacity (typical y 1 to 4 percent) that can be ramped up to regulate frequency. By committing to reserve a portion of capacity in this way, utilities limit their output, losing some production efficiency. Today, batteries are already competitive in the frequency regulation market where they are permitted by regulations, which often require reserve generating capacity to fulfil this role. Batteries wil become more competitive as prices decline. The potential economic impact of energy storage on frequency regulation could be $25 bil ion to $35 bil ion annual y in 2025, net of storage costs. This assumes that battery storage could replace al of the 4 percent of generation capacity set aside for frequency regulation. We have used an average price of frequency regulation of $30 per MWh. Peak load shifting Electricity usage varies by time of day and by season; usual y, the highest demand occurs in the afternoon and early evening, when people come home from work, and during the summer. To meet peak demand (when generation prices are highest), utilities can either build excess generation capacity or purchase electricity from other utilities or from specialized peaker plant suppliers. Energy storage could save costs by enabling utilities to avoid purchasing electricity at peak prices, instead buying (or generating) when it is least expensive, regardless of when it wil be used. The ability to store energy for use at a later time is also useful for integrating energy from renewable sources (wind and solar power, for example) into the electricity supply, due to the intermittent nature of these power sources. The economic impact of using energy storage for peak load shifting could be $10 bil ion to $25 bil ion annual y in 2025. This assumes an LCOE of $55 to $85 per megawatt-hour in 2025 for relevant energy storage solutions and an estimated LCOE of between $25 to $65 per megawatt-hour for on-demand generation (using peaker plants). The general y lower cost of on-demand generation compared with energy storage wil often make peaker plants a more economical choice for utilities. However, at the upper end of the cost range for peak costs,
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 103 energy storage is competitive. Therefore, energy storage could be used for 10 to 20 percent of the roughly 10 to 15 percent of al electricity generation that could be beneficial y shifted. While not included in our estimate of potential economic impact, batteries from electric vehicles could also be used as low-cost energy storage mechanisms for peak load shifting and frequency regulation. Recent studies have found minimal economic benefit to consumers from participating in peak load shifting, but there may stil be an opportunity for providing frequency regulation services.117 Infrastructure deferral Energy storage systems can save money by al owing utilities to delay building transmission and distribution capacity. If peak load wil push a transmission line beyond capacity, energy storage can be placed on the transmission line close to the load source (that is, the area where demand exceeds line capacity) to accommodate peak demand. However, energy storage is financial y viable for this use only in a smal number of cases, such as when lines cannot be upgraded quickly due to long distances, where there are strict permitting requirements due to environmental concerns, or in urban hubs where distribution infrastructure upgrades are exceptional y expensive. Even by 2025, only about 15 percent of electric transmission and distribution infrastructure would be expensive enough to justify investing in storage systems to defer additional investment. As a result of these limitations, the potential economic impact of using energy storage for infrastructure deferral could be approximately $10 bil ion annual y by 2025. BARRIERS AND ENABLERS For the ful economic impact of advanced energy storage to be realized, storage technology wil need to reach cost and performance levels that meet or exceed those of existing alternatives. For electric and hybrid vehicles, for example, this not only means narrowing the gap with conventional ICEs on a cost-of-ownership basis, but also improving responsiveness and driving range between charges. Electric vehicles may also have to become less expensive to purchase and own, since the majority of new car sales in 2025 could be in developing markets. In addition, there wil need to be adequate infrastructure in the form of recharging stations. Governments could facilitate electric and hybrid vehicle adoption through subsidies. In grid applications, there are obstacles to advanced storage options beyond technology cost and performance. In deregulated electricity markets, where generation is separate from transmission and distribution (T&D), some applications for grid energy storage face an uphil battle. Batteries can be used for short-duration load shifting (a generation function) as wel as distribution deferral (a T&D function), but utilities have limited incentives to adopt this solution. Performing each of these services in isolation is less cost competitive than existing solutions; however, when generation and T&D uses are combined, the economic case improves significantly. Regulatory policy is also critical. Regulations can prevent energy storage solutions from competing with generation assets (such as gas-powered peaker plants) for frequency regulation and peak 117 Scott Peterson, Jay Whitacre, and Jay Apt, “The economics of using plug-in hybrid electric vehicle battery packs for grid storage,” Journal of Power Sources, volume 195, number 8, April 2010; Corey D. White and K. Max Zhang, “Using vehicle-to-grid technology for frequency regulation and peak-load reduction,” Journal of Power Sources, volume 196, number 8, 2011.
104 load generation, and prevent batteries from being employed beyond single- use applications. IMPLICATIONS The potential of advancing energy storage technology to power EVs, make electricity more reliable and available in developing nations, and improve power grid efficiency has implications for the auto and energy industries, commercial and residential energy users, and policy makers. For producers of energy storage technology and systems, the coming decade could provide great opportunity; however, for many applications, it wil be up to the industry to make the case for their solutions. Emphasizing the lifetime benefits from investments in energy storage wil be essential for gaining the support of utility company leaders, who tend to invest on decade timescales. Suppliers can also raise the odds for adoption by making energy storage systems work seamlessly with existing grid infrastructure and renewable generation systems, potential y requiring partnerships with companies with core competencies in software, process control systems, and grid integration. To get utilities comfortable with newer, relatively untested battery technology, companies may also want to consider co-investing in initial pilot projects. The onus wil also be on energy storage technology producers to reduce battery costs while improving performance. Governments and research institutions are funding and conducting energy storage research—a potential option for companies looking for cofunding support of this potential y research. Storage solution providers that are not involved in research wil need to keep abreast of how battery component technologies and manufacturing processes are evolving and build flexibility into their manufacturing processes in order to be able to embrace emerging innovations. Vehicle manufacturers wil have opportunities to establish market leadership in providing electric and hybrid vehicles that satisfy consumer expectations regarding performance, utility, safety, convenience, and design. As they do, these manufacturers wil need to plan for multiple scenarios of energy storage technology improvement, improvement of other technologies (for example, advances in ICE technology or the wide adoption of natural gas as a fuel), and oil prices. These scenarios, combined with changing consumer expectations, should determine the pace at which the market embraces electric and hybrid vehicles. Utilities face both risks and opportunities due to advanced energy storage. While energy storage may help improve the quality, reliability, and efficiency of the electricity supply, other uses could affect overal demand, both positively and negatively. As electric and hybrid vehicle adoption accelerates, peak load energy demand could grow by as much as 150 percent if charging is unconstrained (that is, if most drivers come home after work and charge their vehicles when demand is highest). This could place new strain on capacity, requiring new investment in infrastructure. Conversely, rooftop solar and smal -scale wind generation could reduce demand from the grid. In many regions, regulators may require utilities to pay a “green energy premium” for electricity uploaded to the grid from these sources. Alongside these chal enges, utilities could have some important opportunities. With the adoption of EVs, utilities could partner with consumers to use car
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 105 batteries as remote energy storage solutions, effectively using consumers’ plugged in cars as utility storage, especial y for frequency regulation. There is also an opportunity to establish the market for home and road charging bundles, combining energy and hardware at a fee. Final y, to take advantage of distributed generation, utilities could provide long-term leases for storage equipment. Consumers are likely to become increasingly aware of the issues and trade-offs relating to energy storage. As motorists, their decision to buy a hybrid or plug-in vehicle today is not about economics (at least in the absence of tax incentives), but rather about factors such as concern about climate change. In the future, however, electric and hybrid vehicles could become much more cost competitive. As residential utility customers, consumers wil face choices about whether to invest in rooftop solar or smal -scale wind capacity or whether to participate in demand-management schemes. These decisions have implications for the demand for energy storage. Policy makers wil play an important role in determining how much impact energy storage technologies have. Utility regulation should be reviewed to see whether there are incentives or disincentives for investment in grid storage and other relevant applications. The overal goal should be to ensure that energy storage is permitted to compete on an equal footing with other solutions. For example, grid storage should be al owed to compete with generation for frequency regulation and with peaker plants for peak load electricity supply. Introduction of renewable energy quotas could also promote investment in energy storage. Governments can play an important role in supporting energy storage research and development. Many energy storage solutions are stil in the developmental stages and are dependent on investments in basic science research, meaning it can take many years for technologies to be commercialized. Alongside investment in basic science R&D, policy makers may also want to support investment in improving the manufacturing processes of energy storage devices. Advances in energy storage technology could significantly boost the utility and adoption of electric and hybrid vehicles, generating enormous economic impact by 2025. The realization of this potential impact is highly dependent upon retail fuel prices, which are often affected by taxes and other instruments of policy. Policy makers should seriously consider the impact of fuel pricing policies on the future of EVs and careful y weigh the trade-offs. The potential of energy storage for grid applications could be limited by 2025, but this technology nonetheless has longer-term potential to disrupt energy generation and distribution. It is possible to envision a scenario in which distributed renewable generation, combined with cheap storage, and in the presence of higher energy prices, could eventual y lead to significantly increased adoption of distributed renewable power, particularly solar. Eventual y, this could vastly alter the utility industry; the use of oil, gas, coal, and nuclear generation; and the transport industry, ushering in an era of localized energy independence along with drastical y reduced emission levels. Final y, there is little doubt that consumers and many businesses stand to benefit greatly from advances in energy storage technologies, whether they are used to power mobile Internet devices, vehicles, or entire households.
106 9. 3D printing In recent years, 3D printing has attracted increasing attention. The prospect of machines that can print objects much the same way that an inkjet printer creates images on paper has inspired enthusiasts to proclaim that 3D printing wil bring “the next industrial revolution.” Other observers have reacted with skepticism and point to the technology’s current limitations and relatively low level of adoption. Our research leads us to a more nuanced view. While 3D printing does have the potential for disruptive impact on how products are designed, built, distributed, and sold, it could take years before that impact is felt beyond a limited range of goods. Nonetheless, rapidly improving technology and a variety of possible delivery channels for 3D printed goods (such as using the local print shop) could ultimately result in many products being 3D printed. In fact, personal 3D printers are already becoming available for less than $1,000. 3D printing could proliferate rapidly over the coming decade. Sales of personal 3D printers grew 200 to 400 percent every year between 2007 and 2011, and 3D printers are already commonplace for designers, engineers, and architects, who use them to create product designs and prototypes. 3D printing is also gaining traction for direct production of tools, molds, and even final products. These newer uses of 3D printing could enable unprecedented levels of mass customization, shrinking and less-costly supply chains, and even the “democratization” of manufacturing as consumers and entrepreneurs begin to print their own products. Looking longer term, perhaps beyond 2025, one category of 3D printing—bioprinting of living organs—has long-term potential to save or extend many lives. We estimate that 3D printing could generate economic impact of $230 bil ion to $550 bil ion per year by 2025 in the applications we have sized. The largest source of potential impact among sized applications would be from consumer uses, fol owed by direct manufacturing and the use of 3D printing to create tools and molds.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 107 DEFINITION 3D printing belongs to a class of techniques known as additive manufacturing. Additive processes build objects layer-by-layer rather than through molding or subtractive techniques (such as machining). Today, 3D printing can create objects from a variety of materials, including plastic, metal, ceramics, glass, paper, and even living cel s. These materials can come in the form of powders, filaments, liquids, or sheets. With some techniques, a single object can be printed in multiple materials and colors, and a single print job can even produce interconnected moving parts (such as hinges, chain links, or mesh). A variety of 3D printing techniques are in use today, each with its own advantages and drawbacks. Major techniques include selective laser sintering, direct metal laser sintering, fused deposition modeling, stereolithography, and inkjet bioprinting (see Box 9, “Additive manufacturing techniques”). In all cases, objects are formed one layer at a time, each layer on top of the previous, until the final object is complete. With some techniques this is accomplished by melting material and depositing it in layers, while other techniques solidify material in each layer using lasers. In the case of inkjet bioprinting, a combination of scaffolding material and live cel s is sprayed or deposited one tiny dot at a time.118 3D printing has several advantages over conventional construction methods. With 3D printing, an idea can go directly from a file on a designer’s computer to a finished part or product, potential y skipping many traditional manufacturing steps, including procurement of individual parts, creation of parts using molds, machining to carve parts from blocks of material, welding metal parts together, and assembly. 3D printing can also reduce the amount of material wasted in manufacturing, and create objects that are difficult or impossible to produce with traditional techniques, including objects with complex internal structures that add strength, reduce weight, or increase functionality. In metal manufacturing, for example, 3D printing can create objects with an internal honeycomb structure, while bioprinting can create organs with an internal network of blood vessels. Current limitations of 3D printing, which vary by printing technique, include relatively slow build speed, limited object size, limited object detail or resolution, high materials cost, and, in some cases, limited object strength. However, in recent years rapid progress has been made in reducing these limitations. 118 See Savas Tasoglu and Utkan Demirci, “Bioprinting for stem cel research,” Trends in Biotechnology, volume 31, number 1, January 2013.
108 Box 9. Additive manufacturing techniques Selective laser sintering (SLS). In this technique, a layer of powder is deposited on the build platform, and then a laser “draws” a single layer of the object into the powder, fusing the powder together in the right shape. The build platform then moves down and more powder is deposited to draw the next layer. SLS does not require any supporting structure, which makes it capable of producing very complex parts. SLS has been used mostly to create prototypes but recently has become practical for limited-run manufacturing. General Electric, for example, recently bought an SLS engineering company to build parts for its new short-haul commercial jet engine. Direct metal laser sintering (DMLS). DMLS is similar to selective laser sintering but deposits completely melted metal powder free of binder or fluxing agent, thus building a part with al of the desirable properties of the original metal material. DMLS is used for rapid tooling development, medical implants, and aerospace parts for high-heat applications. Fused deposition modeling (FDM). A filament of plastic resin, wax, or another material is extruded through a heated nozzle in a process in which each layer of the part is traced on top of the previous layer. If a supporting structure is required, the system uses a second nozzle to build that structure from a material that is later discarded (such as polyvinyl alcohol). FDM is mainly used for single- and multipart prototyping and low-volume manufacturing of parts, including structural components. Stereolithography (SLA). A laser or other UV light source is aimed onto the surface of a pool of photopolymer (light-sensitive resin). The laser draws a single layer on the liquid surface; the build platform then moves down, and more fluid is released to draw the next layer. SLA is widely used for rapid prototyping and for creating intricate shapes with high- quality finishes, such as jewelry. Laminated Object Manufacturing (LOM). A sheet of material (paper, plastic, or metal) is fed over the build platform, adhered to the layer below by a heated rol er, and a laser cuts the outline of the part in the current layer. LOM is typical y used for form/fit testing, rapid tooling patterns, and producing less detailed parts, potential y in full color. Inkjet-bioprinting. Bioprinting uses a technique similar to that of inkjet printers, in which a precisely positioned nozzle deposits one tiny dot of ink at a time to form shapes. In the case of bioprinting, the material used is human cel s rather than ink. The object is built by spraying a combination of scaffolding material (such as sugar-based hydrogel) and living cel s grown from a patient’s own tissues. After printing, the tissue is placed in a chamber with the right temperature and oxygen conditions to facilitate cel growth. When the cel s have combined, the scaffolding material is removed and the tissue is ready to be transplanted.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 109 POTENTIAL FOR ACCELERATION Until now, 3D printing has been used primarily for rapid prototyping. However, the technology has been evolving since the 1980s, and it may have reached a tipping point at which much more widespread adoption could occur. Advances include improvement in the performance of additive manufacturing machinery, an expanding range of possible materials, and fal ing prices for both printers and materials. Importantly, major initial patents have expired or wil expire soon. When the original patent for fused deposition modeling (FDM) expired in 2009, 3D printing systems created through an open source project cal ed RepRap (for replicating rapid prototyper) started to become commercial y available. This has helped to spread free software and encourage rapid innovation for 3D printing using a variation on the FDM process. The last of the original selective laser sintering patents filed by the process’s inventors (Carl Deckard and Joe Beaman) wil expire in June 2014, which could spawn another round of innovation and low- cost commercialization. 3D printing for producing complex, low volume, and highly customizable products is already accelerating. Boeing currently prints 200 different parts for ten aircraft platforms. In health care, manufacturers have been offering printed custom hearing aid earpieces, sel ing more than one mil ion units in 2011. In addition, more than 40,000 acetabular hip cups (the socket for hip joint replacements) have been built using 3D printing. The dental appliance maker Invisalign produces 50,000 to 60,000 appliances per day using stereolithography printers. Entrepreneurs are also designing and sel ing 3D printed products, such as iPhone cases, often using services like Sculpteo and Shapeways that al ow designers to upload designs that the company then prints and ships to consumers. As 3D printing continues to mature and grow, it has the potential to address many important needs. In intensely competitive consumer products markets, 3D printing can meet rising expectations for quality design and personalization, including better fit for items such as shoes or helmets (see Box 10, “The promise of 3D printing: Everything made to order in one step”). 3D printing also has the potential to address concerns about the waste and environmental impact of traditional manufacturing processes and supply chains. Direct product manufacturing using printing can reduce the number of steps required for parts production, transportation, assembly, and distribution, and it can reduce the amount of material wasted in comparison with subtractive methods. In medicine, the ability to print body parts from the patient’s own cel s could improve transplant success rates and prevent deaths that occur due to patients having to wait for donor organs. Improvements in speed and performance and fal ing costs wil likely accelerate the spread of 3D printing in the coming decade. The average industrial printer now sel s for about $75,000, and some machines cost more than $1 mil ion. However, these costs are widely expected to decline rapidly in coming years as production volumes grow. Advances are also under way that could dramatical y improve the output speed and quality of 3D printers. For example, recent work at the Fraunhofer Institute for Laser Technology points to the potential for a fourfold increase in printing speeds for metal objects. On the consumer side, prices for basic 3D printers using fused deposition modeling technology have declined from $30,000 a few years ago to less than $1,000 for some models. Unit sales of consumer 3D printers remain smal , with
110 about 23,000 printers sold in 2011, but these sales are growing rapidly, with more than 300 percent in average annual growth between 2007 and 2011. Meanwhile, 3D printing services are spreading rapidly. Shapeways already has more than 8,000 online shops and shipped 1 mil ion parts in 2012. In late 2012, Staples announced plans to rol out a new 3D printing service in the Netherlands and Belgium that wil al ow customers to upload 3D designs and pick up the finished items at their local Staples store.119 The materials used in 3D printing stil remain costly (general y about 50 to 100 times greater than materials used for injection molding), but prices are declining rapidly and can be expected to decline further as volumes increase. Chinese plastics suppliers have already started to sel plastic filaments at very competitive prices—just five times the price of production grade plastics—and new extruders have been developed that can turn production-grade plastics into filaments. In addition, new types of materials are being adapted for additive manufacturing every year. Some newer polymer types that can work in 3D printers offer flexibility, electrical conductivity, and even biocompatibility (e.g., for implants). An important step could be the development of standardized materials that can be used by 3D printing systems from different vendors; today, each manufacturer requires its own certified materials. Box 10. Vision: Everything made to order in one step Imagine that you need new shoes. Instead of going to the store, you go online and buy a cool shoe design for a few dol ars, or just download one for free. With a few clicks, you use a mobile phone app to scan your foot, select a few colors, and upload the design to print at a local 3D printing shop. Next you log on to a popular furniture site, browse for a while, and select an interesting metal and plastic chair, specifying a few adjustments to the size. A few hours later, on the way to the gym, you visit the 3D printing shop and try on your new shoes, which fit perfectly. Your new 3D-printed chair is delivered to your doorstep later that week. On the way to work, one of those new passenger planes passes overhead. In its ads, the airline promises to get you there faster and with lower carbon emissions than with the older planes in its fleet. They can make this boast because of the way 3D printing has changed aerospace products. The strength and durability of airplane parts has increased. The jet engine block is made from a single piece of metal with an internal structure that removes weight with little loss of strength. To top it off, these 3D printed parts are made using far fewer resources and with almost no waste. You also think about your father, who needed a knee joint replacement last year. Amazingly, the doctors were able to scan his leg and 3D print a titanium replacement joint that fit perfectly. They even bioprinted replacement ligaments for the knee using a sample of his own cel s, and after rehabilitation, he’s back to playing basketball with his grandkids. 119 Objects printed using laminating object manufacturing and paper material have similar material characteristics to wood and can be worked in a similar way (e.g., by sanding, carving, and dril ing).
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 111 POTENTIAL ECONOMIC IMPACT BY 2025 We estimate that 3D printing could generate economic impact of $230 bil ion to $550 bil ion per year across our sized applications by 2025 (Exhibit 11). The largest source of potential impact among the sized applications would come from consumer uses, fol owed by direct manufacturing (i.e., using 3D printing to produce finished goods) and using 3D printing to make molds.
Exhibit 11 Sized applications of 3D printing could have direct economic impact of $230 billion to $550 billion per year in 2025 Potential economic impact of sized Sized applications in 2025 Estimated scope Estimated potential Potential productivity or applications $ bil ion, annually in 2025 reach in 2025 value gains in 2025 $4 tril ion in sales of 5–10% of relevant 60–80% value increase consumer products products (e.g., toys) per 3D-printed product that might be could be 3D printable, – 35–60% cost savings Consumer use 100– 3D printed assuming easy to consumers of 3D printing 300 consumer access – 10% added value from customization
Direct product 100– $300 bil ion spending 30–50% of products 40–55% cost savings to manufacturing1 200 on complex, low- in relevant categories buyers of 3D-printed
volume items such as replaceable with 3D products
implants and tools printing Tool and mold 30– $470 bil ion spending manufacturing 50 on complex, low- volume parts in transportation Other potential $360 bil ion global 30–50% of injection- 30% production cost applications market for injection- molded plastics reduction using superior (not sized) molded plastics produced with 3D- 3D-printed molds Sum of sized printed molds potential 230– economic 550 impacts 1 Focuses on use of 3D printing to directly manufacture low-volume, high-value parts in the medical and transport manufacturing industries. Other potential y impactful applications might include manufacturing of low-volume, high-value replacement parts for other industries. NOTE: Estimates of potential economic impact are for some applications only and are not comprehensive estimates of total potential impact. Estimates include consumer surplus and cannot be related to potential company revenue, market size, or GDP impact. We do not size possible surplus shifts among companies and industries, or between companies and consumers. These estimates are not risk- or probability-adjusted. Numbers may not sum due to rounding. SOURCE: McKinsey Global Institute analysis Consumer uses We estimate that consumer use of 3D printing could have potential economic impact of $100 bil ion to $300 bil ion per year by 2025, based on reduced cost (compared with buying items through retailers) and the value of customization. 3D printing could have meaningful impact on certain consumer product categories, including toys, accessories, jewelry, footwear, ceramics, and simple apparel. These products are relatively easy to make using 3D printing technology and could have high customization value for consumers. Global sales of products in these categories could grow to $4 tril ion a year (at retail prices) by 2025. It is possible that most, if not al , consumers of these products could have access to 3D printing by 2025, whether by owning a 3D printer, using a 3D printer in a local store, or ordering 3D printed products online. We estimate that consumers might 3D print 5 to 10 percent of these products by 2025, based on the products’ material composition, complexity, cost, and the potential convenience and enjoyment of printing compared with buying for consumers. A potential 35 to 60 percent cost savings is possible for consumers self-printing these goods despite higher material costs (the materials required for the products we focus on here, primarily plastics, are relatively inexpensive and 11
112 getting cheaper). The savings over retail come not only from eliminating the costs of wholesale and retail distribution, but also from reducing the costs of design and advertising embedded in the price of products. It is possible that consumers wil pay for 3D printing designs, but it is also possible that many free designs wil be available online. Final y, customization might be worth 10 percent or even more of the value for some 3D printed consumer products.120 Direct manufacturing Even in 2025, traditional manufacturing techniques wil almost certainly have a large cost advantage over additive manufacturing for most high-volume products. However, 3D printing could become an increasingly common approach for highly complex, low-volume, highly customizable parts. If used in this way, we estimate that 3D printing could generate $100 bil ion to $200 bil ion in economic impact per year by 2025 from direct manufacturing of parts. The market for complex, low-volume, highly customizable parts, such as medical implants and engine components, could be $770 bil ion annual y by 2025, and it is possible that some 30 to 50 percent of these products could be 3D printed. These products could cost 40 to 55 percent less due to the elimination of tooling costs, reduction in wasted material, and reduced handling costs. Tool and mold manufacturing Even by 2025, the large majority of parts and products wil stil be manufactured more efficiently with techniques such as injection molding. 3D printing, however, has the potential to create significant value by shortening setup times, eliminating tooling errors, and producing molds that can actual y increase the productivity of the injection molding process. For example, 3D printed molds can more easily include “conformal” cooling channels, which al ow for more rapid cooling, significantly reducing cycle times and improving part quality. We estimate that 3D printing of tools and molds could generate $30 bil ion to $50 bil ion in economic impact per year by 2025, based on an estimated $360 bil ion cost base for production of injection molded plastics in 2025 and assuming that about 30 to 50 percent of these plastics could be produced with 3D printed molds at around 30 percent less cost. BARRIERS AND ENABLERS Despite improvements in 3D printing technology, remaining limitations, particularly material costs and build speeds, could constrain wide-scale adoption. However, both materials costs and speed could improve dramatical y by 2025 for many techniques based on the current evolution of the technology. If materials costs and build speeds fail to improve significantly, it could substantial y lessen the economic impact of 3D printing. Much of the potential value of 3D printing for consumers and entrepreneurs wil depend on the emergence of an “ecosystem” to support users. Online 3D object exchanges like thingiverse point to a potential future in which object designs are widely exchanged and purchased like music files, greatly facilitating the spread of 3D printing adoption. The success of 3D printing also depends on improvements in products such as design software, 3D scanners, and supporting software applications and tools. Commercial 3D scanners are an important 120 For example, Nike currently offers customizable NikeID shoes at a surcharge of approximately 30 percent over the price of standard designs of similar quality.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 113 enabling technology. Hobbyists are currently using Microsoft’s Xbox Kinect to create 3D scans, and smartphones can be converted into basic 3D scanners via the use of an app. Consumer-oriented, sub-$1,000 3D scanners could soon be coming to market as wel . The expiration of key patents for printing technologies could inspire waves of low-cost, highly capable 3D printers for businesses and consumers. IMPLICATIONS While the range of products that consumers and companies choose to 3D print could be limited at first, the ability to easily design and self-manufacture products could create significant consumer surplus and even influence consumer culture. Access to 3D printers is already inspiring a “maker” subculture in which enthusiasts share designs and ideas. For example, the 3D printing service Shapeways already has more than 10,000 crowd-sourced 3D models for jewelry and other items that have been uploaded by consumers. 3D printing could eventual y spawn the same kind of dynamic, complex ecosystem that exists in software and Web development—in which developers can easily share and col aborate with one another—extending this kind of innovation ecosystem to the creation of physical objects. Budding product designers and entrepreneurs can use 3D printing to quickly reach a mass, even global, audience. Makers of 3D printing devices and service providers should consider how best to stake out the most favorable positions within this ecosystem, whether by establishing a brand for consumer 3D printers, establishing a marketplace for 3D designs, or opening go-to 3D print shops (either online or in brick-and-mortar locations). Ultimately, a gradual y rising share of sales in categories such as toys and personal accessories could shift to either home production or 3D printing centers. Leaders in these categories should identify how to add value for consumers in ways that home-printed products cannot. Manufacturers of consumer products that could become 3D-printable should consider new ways to customize products to match some of the advantages of home printing. At the same time, they should fol ow closely the evolution of online exchanges for 3D printable designs and careful y manage their intel ectual property rights while proactively leveraging these exchanges to distribute their products. Internal y, these manufacturers should take advantage of 3D printing for rapid prototyping to speed designs to market and keep careful watch for the moment when improvements in technology might make 3D printing an economical y viable production method for them. Materials, manufacturing, and logistics value chains related to products that are candidates for 3D printing could also be affected by this technology. Direct production of goods by consumers could affect demand for some materials and global shipping volumes (although only in a smal way initial y). Leaders of businesses that could be affected by this new production ecosystem should think about what role they want to play and develop strategies to compete. For example, Staples’ decision to experiment with 3D printing services suggests that there may be opportunities to offer other services and products to entrepreneurs and consumers who use 3D printing.
114 Access to 3D printing could actual y make some manufacturing sectors more competitive. For industries with high-value goods, in which rapid innovation is more important than absolute cost, the combination of 3D printing of products and advanced robotics (see Chapter 5, “Advanced robotics”) could make proximity to end markets and access to highly skil ed talent more important than hourly labor rates in determining where production is located. This could lead some advanced economy companies to produce more goods domestical y, boosting local economies. However, this may not create many manufacturing jobs, as the 3D printing process is highly automated. 3D printing poses opportunities as wel as chal enges for policy makers in both advanced and developing economies. Societies can benefit from products that are made with less waste that do not require transport over great distances and, therefore, have less impact on the environment.121 Policy makers should consider supporting the development of 3D printing, in particular by funding research in 3D printing technologies. The chal enges for policy makers include addressing regulatory issues—such as approving new materials for use—ensuring appropriate intel ectual property protections, and assigning legal liability for problems and accidents caused by 3D printed products. Governments wil also be cal ed upon to clarify how intel ectual property rights wil be protected. 3D printers have already been used to make handguns, raising another set of issues. Policy makers face the chal enge of evaluating and addressing these risks without stifling innovation or limiting the value that this technology can provide. 121 According to the US Department of Energy, 3D printing can significantly reduce energy and materials use compared with traditional manufacturing methods. See www1.eere.energy.gov/ manufacturing/pdfs/additive_manufacturing.pdf.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 115 10. Advanced materials History shows that breakthroughs in the quality or cost of basic materials can drive cycles of disruptive growth. For example, the so-cal ed Second Industrial Revolution, a period of rapid innovation and expansion from about 1860 to 1920, was enabled in part by advances in steel manufacturing technologies that provided a lighter, stronger, and more versatile material that went into thousands of mass-produced products, as wel as into bridges, skyscrapers, and ships. A new revolution in materials has been taking shape in research laboratories around the world during the past few decades. Scientists are perfecting new ways to manipulate matter to produce advanced materials with unheard-of attributes that could enable innovations in fields ranging from infrastructure construction to medicine. These advances include so-cal ed smart materials that are self-healing or self-cleaning, memory metals that can revert to their original shapes, piezoelectric ceramics and crystals that turn pressure into energy, and nanomaterials. While many of these advanced materials may have interesting and potential y high-impact applications, it remains far from clear whether most wil be capable of driving significant impact by 2025. We therefore focus here on advanced nanomaterials, which could be used to create many other smart materials (including those mentioned above) and for a host of other important applications (see Box 11, “The vision: A new dimension”). Nanomaterials are made possible by manipulating matter at nanoscale (less than 100 nanometers, or approximately molecular scale). At nanoscale, ordinary substances like carbon and clay take on surprising properties—including greater reactivity, unusual electrical properties, and enormous strength per unit of weight—that can enable the creation of new types of medicine, super-slick coatings, and stronger composites. Invisible to the naked eye, nanomaterials have already found their way into products as varied as pharmaceuticals, sunscreens, bacteria-killing socks, and composite bicycle frames. And recent generations of semiconductors have features in the tens of nanometers, in effect making them nanotechnology. Nanomaterials could have wide applications across health care, electronics, composites, solar cel s, water desalination and filtration, chemicals, and catalysts. However, producing the nanomaterials required for many of these applications remains extremely expensive. The nanomaterials in use today are mostly particles (silver, clay, and metal oxides) that are relatively simple and easy to produce.122 122 Relatively simple nanoparticles are already being used in products such as car bumpers, scratch-resistant coatings, bacteria-kil ing socks, and sunscreen. This has led some to describe nanotechnology as a tril ion-dol ar industry, referring to the total revenue of products that incorporate some form of nanomaterials. See, Mihail C. Roco and Wil iam Sims Bainbridge, Societal implications of nanoscience and nanotechnology, National Science Foundation, March 2001.
116 Box 11. Vision: A new dimension Imagine a world in which advanced nanomaterials have revolutionized medical diagnostics and treatment. Using a simple blood test and inexpensive gene sequencing (see Chapter 6, “Next-generation genomics”), patients would be checked for illnesses, including various cancers. When cancer is detected, a customized dose of cancer-kil ing chemicals would be attached to nanoparticles that would deliver the treatment to the cancer cel s without affecting healthy cel s. Nano-based drug delivery could make chemotherapy more effective, helping to save lives while reducing the side effects. Advanced nanomaterials could someday be used to manufacture al sorts of goods. Consumer electronic devices would have unprecedented levels of speed and power because their circuitry would be based on highly conductive graphene. With graphene-based supercapacitor batteries, it would take only a few minutes to give a mobile Internet device a charge that would last for a week. And thanks to nanocomposites, your self-driving electric car can go 300 miles without recharging and is stil lighter and stronger than old-fashioned steel and plastic vehicles. Your tablet computer would no longer be a stiff, clunky, book-shaped device, but rather a thin sheet that can be rol ed up and put in a pocket. In homes and offices, wal s are covered with ultra-thin nanomaterial-based displays that can be used to view information, enjoy entertainment, or simply to place a pretty picture on the wal . And there would be no need to worry about wasting electricity: these displays consume low levels of power and the highly efficient, graphene-based solar cel s and energy storage systems in homes would supply most consumer power needs. Nanomaterials may also hold the key to solving the world’s water shortage issues. Graphene-based filters could both turn salt water into freshwater and remove al impurities, making water shortages a thing of the past.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 117 More advanced nanomaterials such as graphene (ultrathin sheets of graphite) and carbon nanotubes (tubular graphene) could eventual y be used to create superefficient batteries; thin, flexible, energy efficient displays; ultralight, superstrong structural materials; and even the next generation of semiconductor chips. However, producing graphene and nanotubes in large quantities remains prohibitively expensive (as much as $700 per gram for carbon nanotubes) and is expected to remain so for years. Over the coming decade, the most important application of advanced nanomaterials could be the use of nanoparticles to create new targeted treatments for cancer. We estimate that the use of advanced nanomaterials for targeted drug delivery for cancer alone could generate an economic impact of $150 bil ion to $500 bil ion annual y by 2025. After many years of delivering more promise than visible progress, nanotechnology is often viewed as overhyped. The truth is that nanotechnology, albeit in its more basic, invisible forms, is already a reality today and wil have a growing role in industry, medicine, and the lives of consumers in years to come. Over the coming decade, the ful potential of advanced nanomaterials may only begin to be felt, but these materials wil likely continue to attract considerable interest and R&D investment. Nanomaterials could begin to open up major opportunities for health-care technology companies, pharmaceutical companies, and health-care providers. Business leaders, particularly in health care, manufacturing, and electronics, should consider now how these materials could be used to create new products or make existing products better, and invest accordingly. Meanwhile, policy makers wil need to address unanswered questions regarding the safety of nanomaterials. DEFINITION Any use or manipulation of materials with features at a scale of less than 100 nanometers (roughly molecular scale) can qualify as nanotechnology. This is a very, very smal scale indeed. Each nanometer is one bil ionth of a meter; the width of a human hair is typical y 20,000 to 80,000 nanometers. Nanoscale objects and machines can be created in a variety of ways, including direct manipulation of molecule-sized nanoparticles using tools such as atomic-force microscopes, or electron beam or laser lithography capable of printing two- or three-dimensional structures with nanoscale features.123 Since complex nanoscale structures such as nanomachines (including nano electromechanical machines, or NEMS) are currently experimental and very difficult to construct, and methods for large-scale production may not be developed for another decade, we focus in this chapter on nearer-term applications of advanced nanomaterials. Nanomaterials can have amazing properties. At nanoscale, science enters the strange realm of quantum mechanics. Nanoparticles, for example, general y have far greater surface area per unit of volume (up to 2,000 square meters per gram) than other materials and are thus highly reactive (and bio-reactive), making them useful in medicine. Nanoscale materials can also have unusual electromagnetic, thermal, and optical characteristics, which could enable advances across many of the technologies described in other chapters of this report, including the next- generation sensors and actuators use in advanced robotics and in Internet of Things applications (see Chapters 5 and 3). The properties of nanomaterials could make them useful in creating a variety of other advanced materials (see Box 12, “Living in a materials world”). 123 “3D-printer with nano-precision,” Phys, http://phys.org/news/2012-03-3d-printer-nano- precision.html.
118 Box 12. Living in a materials world While nanomaterials offer considerable promise, particularly long term, they are by no means the only type of advanced materials that could drive economic impact over the coming decades. Below are some other types of materials that could help build a better, cleaner world, though each has its own engineering, production, and performance chal enges to overcome. “Green” materials attempt to solve environmental issues. Low-CO 2 concrete, for example, could reduce emissions from concrete production, which are estimated to account for 5 percent of total CO emissions.1 2 Adding advanced materials reduces the amount of fuel required to burn and grind the ingredients of concrete and reduce the need to de- carbonate limestone within the kiln. Self-healing materials take their inspiration from biological systems that can self-organize and self-repair. Self-healing materials would reduce the need for costly maintenance by healing themselves when damage occurs. One example is self-healing concrete, which would include ingredients that are automatical y released or that expand to fil cracks when they appear. Piezoelectric materials that turn pressure into electricity are not new, but researchers continue to find new potential applications, such as generating electrical energy from movement. Eventual y, it could be possible to capture electricity from the movement of pedestrians to generate electricity or to incorporate piezoelectric materials into clothing to power mobile Internet devices. Memory metals revert to a prior shape when heated to a specific temperature. These materials are being considered as a way of producing movement in light, inexpensive robots—using a charge to expand or contract the material, imitating muscle movement. Some versions of memory metals can even be “programmed” to take on multiple shapes at different temperatures. Advanced composites could help build strong, lighter components for vehicles, including aircraft. In addition to next-generation nanocomposites, ongoing advances in composites made from carbon fiber and other materials could make it possible to substitute composites for materials such as aluminum in more applications. These advances include new ways of producing and binding carbon fiber, al owing for less expensive fabrication. 1 David Bradley, “10 Technology Breakthroughs,” MIT Technology Review, 2010.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 119 Today, simple nanoparticles are the most widely used nanomaterials. They are used in coatings, paints, sensors, chemical catalysts, and food packaging. Silver nanoparticles, which have antimicrobial properties, have been added to laundry detergents and even woven into socks. Zinc oxide nanoparticles have interesting optical properties (al owing some wavelengths through while blocking others), making them useful ingredients in some sunscreens. Clay nanoparticles can make lighter, stronger, and more elastic composites, in part by reducing the amount of fil er material required and increasing binding, and are used in some car bumpers. Clay nanoparticles also improve barrier properties and are therefore used in some plastic food packaging to increase shelf life. In the near term, medicine could be the most promising area for adoption of advanced nanomaterials. The large surface area and high reactivity of many nanomaterials could make them powerful diagnostic tools for many diseases, including cancer. Nanoparticles can also be used to create potential y lifesaving medicines that can target specific tissues or cel s. This can create therapies that are more effective and reduce harmful side effects. For example, researchers are working on ways to use gold and silver nanoparticles, as wel as liposomes (nano-sized bubbles made from the material of cel membranes) for targeted drug delivery. Nanomaterials can be combined with cancer-kil ing substances and then delivered to a tumor in a more precise way than current options, reducing the damage to healthy cel s and other side effects of conventional chemotherapy. In late 2012, AstraZeneca announced that it is developing a treatment that uses gold nanoparticles to convey the cancer-kil ing drug TNF (tumor necrosis factor) to specific cancer sites. TNF is normal y highly toxic, but might be safely delivered using nanoparticles because it would be targeted directly to tumors.124 Nanoparticles can be used to target cancer cel s passively (by taking advantage of these cel s’ increased tendency to absorb such particles relative to normal cel s) or actively (by attaching molecules designed to specifical y seek out or bind to cancer cel s, such as peptides).125 Graphene, which is composed of one-atom-thick sheets of carbon hexagons, is being produced today, but only in limited quantities and at high cost. When this material can be mass-produced cost-effectively, its impact could be quite disruptive. Graphene is one-sixth the weight of steel per unit of volume but more than 100 times as strong. Graphene can be compressed without fracturing, recovering its original shape after being pressurized to more than 3,000 atmospheres. Graphene also has 35 percent less electrical resistance than copper and ten times the conductivity of copper and aluminum, making it an excel ent material for building electrical circuits; it has been estimated that graphene could yield terahertz processor speeds (about 1,000 times faster than today’s fastest microchips) and could one day replace silicon entirely. In 2011 IBM created the first integrated circuit based on a graphene transistor.126 However, integrating graphene into chips at scale has so far proven chal enging. 124 Karen Weintraub, “Gold particles could deliver cancer drugs,” Boston Globe, December 24, 2012. 125 Peptides are biological molecules made of amino acids that can perform many functions, including kil ing cel s or selectively delivering drugs to specific types of cel s or tissues. 126 Yu-Ming Lin et al., “Wafer-scale graphene integrated circuit,” June 2011.
120 Graphene and carbon nanotubes have a host of potential applications. Because of their unique chemical and electrical properties, including large surface area and high reactivity, carbon nanotubes could act as extremely powerful sensors, al owing the detection of trace molecules of dangerous substances or biomarkers for diseases such as cancer. Graphene-based supercapacitors are being developed with the goal of producing ultra-efficient batteries that could charge in seconds yet power a smartphone or other device for days.127 Graphene could also potential y be used to create highly efficient solar cel s (see Chapter 12, “Renewable energy”), or as a coating in lithium-ion battery electrodes, enabling faster charging and greater storage capacity, a potential boost to the adoption of electric vehicles (see Chapter 8, “Energy storage”). And both graphene and carbon nanotubes can be used as electron emitters to build anodes for highly efficient, super thin, and (with graphene) possibly flexible and transparent displays. Final y, because of its unique absorptive qualities, graphene may help improve access to potable water, a growing issue in many parts of the world. Lockheed Martin recently announced progress in creating graphene-based filters that could produce drinking water from sea water at a smal fraction of the cost of current methods, such as reverse osmosis.128 However, even as potential applications multiply, it remains unclear whether graphene and carbon nanotube production and handling processes can be scaled up cost-effectively. Perfecting scalable, cost-effective production techniques could wel take more than a decade. Graphene and carbon nanotube prices vary widely based on purity, size, form, and (for graphene) substrate material. Today, the sel ing price for 50 mil imeter x 5 mil imeter monolayer graphene thin films manufactured by the company Graphene Square ranges from $264 to $819. Graphene nanoplatelets (five to eight nanometers thick) manufactured by XG Sciences are sold for about $220 to $230 per kilogram. Prices for carbon nanotubes range from $50 per gram to more than $700. Another promising nanomaterial is quantum dots—nanoparticle semiconductors with unique optical properties. Quantum dots can efficiently produce colored light, potential y making them useful in electronic displays. They could also be used as medical diagnostic tools in place of traditional organic dyes, targeting tumors and lighting up under imaging (if toxicity risks can be addressed). Quantum dots are also a possible candidate for creating qbits (quantum bits), the informational unit for quantum computers.129 Quantum computers could, in effect, perform many operations simultaneously by exploiting quantum mechanics. POTENTIAL FOR ACCELERATION In the coming decade, it is possible that nano-based materials and processes could help meet needs in medicine and perhaps see adoption in electronic products, such as displays. Medical diagnostics and treatments enabled by graphene, carbon nanotubes, quantum dots, gold nanoparticles, or biological nanomaterials such as liposomes and peptides could save and extend many lives. There is ever-growing demand for better portable electronic devices and displays (see Chapter 1, “Mobile Internet”). Nanotubes, graphene, and quantum dots could 127 Maher El-Kady and Richard Kaner, “Scalable fabrication of high-power graphene micro- supercapacitors for flexible and on-chip energy storage,” Nature Communications, volume 4, February 2013. 128 Company press release, March 18, 2013. 129 T. D. Ladd et al., “Quantum computers,” Nature, volume 464, number 7285, March 2010.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 121 help make these displays brighter, thinner, and more energy efficient, possibly even al owing these displays to be made transparent and flexible—capable of being “papered” onto wal s, windows, or curved surfaces. However, it is far from clear whether the use of nanomaterials for such displays wil be economical by 2025, given their high cost and the availability of substitute technologies, such as LCD and OLED. Research is bringing the technology for large-scale manufacturing of graphene and nanotubes nearer. In 2012, researchers at Universiti Sains Malaysia announced advances in creating carbon nanotubes that they claim could reduce the price of this material to $15 to $35 per gram.130 In 2011 Lockheed Martin announced that its F-35 fighter jet wil use carbon nanotube composite plastics in some structural parts thanks to an improved manufacturing process, which the company says reduces the cost of producing nanotube-based parts by 90 percent.131 Samsung and IBM are funding R&D in commercial applications of graphene and nanotubes, and several major research institutions are also investigating graphene. Some of the biggest chal enges this research addresses include manufacturing large sheets of graphene and long strands of nanotubes. Another concern is the possible health effects of loose nanomaterials. There is evidence that nanotubes, when inhaled, can have damaging health effects similar to asbestos, though the risk of inhalation could be low for many applications.132 Despite the difficulties of large-scale production, it is expected that the global market for graphene wil grow rapidly in the coming decade, though estimates range widely. According to BCC Research, the global market for commercial products made using graphene is negligible today, but the company predicts that it wil grow to $123 mil ion by 2017 and to $987 mil ion by 2022, with the largest markets being for capacitors, structural materials, and computing.133 BCC also estimates that the global market for carbon nanotubes could reach $527 mil ion by 2016.134 By comparison, world aluminum production in 2012 was 45 mil ion tons, for a total market value of $112 bil ion.135 POTENTIAL ECONOMIC IMPACT BY 2025 For the purposes of estimating the potential economic impact of advanced nanomaterials, we focus on applications in medicine, specifical y drug delivery for cancer patients. The application of advanced nanomaterials for medical purposes has relatively high potential by 2025 given the types of advanced nanomaterials likely to be used, the limited quantities needed, the maturity of the production processes for these materials, and the high wil ingness of consumers 130 Jasmine Leong, “New method for continuous production of carbon nanotubes by USM researchers,” Scientific Malaysian, April 23, 2012. 131 Stephen Trimble, “Lockheed Martin reveals F-35 to feature nanocomposite structures,” Flight International, May 26, 2011. 132 Paul Borm and Wolfgang Kreyling, “Toxicological hazards of inhaled nanoparticles—Potential implications for drug delivery,” Journal of Nanoscience and Nanotechnology, volume 4, number 5, October 2004. 133 Andrew McWil iams, Graphene: Technologies, applications, and markets, BCC Research, July 2012. 134 John Oliver, Global markets and technologies for carbon nanotubes, BCC Research, July 2012. 135 The aluminium market analysis, financials and forecasting 2012–2017, ASDReports, May 2012.
122 to pay for potential y life-saving treatments. For this application, we estimate that advanced nanomaterials could have a potential economic impact of $150 bil ion to $500 bil ion annually by 2025 (Exhibit 12).
Exhibit 12 Sized applications of advanced materials could have direct economic impact of $150 billion to $500 billion per year in 2025 Potential economic impact of sized Sized applications in 2025 Estimated scope Estimated potential Potential productivity or applications $ bil ion, annually in 2025 reach in 2025 value gains in 2025 20 mil ion new 5–10% of cancer $130,000–230,000 QALY value cancer cases patients could benefit created per patient1 Drug 150– worldwide in 2025 from nano-based drug $100,000–200,000 for delivery 500 delivery treatments 1–2 years increased life expectancy $30,000 from reduced chemotherapy side effects Other potential Example applications not sized include nanomaterials for electronics and
applications composites and applications of other advanced and smart materials, such as self-
(not sized) healing concrete or memory metals
Sum of sized potential 150– economic 500 impacts 1 QALY is quality-adjusted life year. NOTE: Estimates of potential economic impact are for some applications only and are not comprehensive estimates of total potential impact. Estimates include consumer surplus and cannot be related to potential company revenue, market size, or GDP impact. We do not size possible surplus shifts among companies and industries, or between companies and consumers. These estimates are not risk- or probability-adjusted. SOURCE: McKinsey Global Institute analysis Other uses may be developed for nanomaterials in electronics, solar cel s, composites, and water systems during this decade; however, their potential economic impact could be limited through 2025 due to the engineering and production chal enges involved, the cost of the nanomaterials required, and the cost and effectiveness of existing or expected substitutes. We also do not size the potential impact from nano-based therapies that seem to be farther from potential use than cancer treatments (for example, HIV treatment). Drug delivery for cancer patients Our estimate of potential impact of $150 bil ion to $500 bil ion annual y by 2025 is based on the assumption that 5 to 10 percent of the 20 mil ion cancer patients that we estimate could exist in 2025 could be treated with nano-based therapies. Conclusive data on the relative efficacy of such treatments remains limited given the newness of this technology. However, based on current research, it seems possible that future nano-based drug delivery treatments could significantly improve outcomes and reduce side effects for many cancer patients. Here we assume that nano-based treatments could add one to two years of life expectancy and significantly improve quality of life during one year of treatment. We estimate the value of these impacts to be $130,000 to $230,000 per patient.136 136 We use a quality of adjusted life year (QALY) value of $100,000 and assume one to two years of extended life; we value the improved quality of life for patients undergoing treatment at $30,000 per patient, assuming an a 12 verage 30 perce nt quality of life improvement during one year of treatment.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 123 Significant R&D spending is already going into the development of nanomaterials for drug delivery, particularly liposomes and gold nanoparticles. Celgene, a pharmaceuticals company, is seeking approval for a pancreatic cancer therapy that would use nanoparticles to deliver the chemotherapy drug paclitaxel to targeted cel s. Celgene has said that it expects sales of $2 bil ion a year by 2017 for this treatment, which would make it the first nano-blockbuster.137 According to data from PharmaProjects, as of March 2013 there were 80 nano-based drugs in the development pipeline, with seven in phase 3 clinical trials. Of the 80 drugs in the pipeline, 43 were for cancer treatments. The nanotechnology-enabled drug delivery market is expected to grow to $136 bil ion by 2021, with liposomes and gold nanoparticles accounting for 45 percent of this market, according to market researcher Cientifica.138 BARRIERS AND ENABLERS The use of advanced nanomaterials in medicine could drive significant economic impact by 2025, but realizing this potential depends on whether specific nano- based drugs can be successful y developed and approved at reasonable cost. Drug development of any kind is general y very costly, and the vast majority of new drug candidates are never approved. However, given the number of nano- based drugs in various stages of development, it is possible that a number of them wil come to market within several years. Unfortunately, many of these drugs may be very expensive. Abraxane, Celgene’s nano-based drug, wil reportedly cost pancreatic cancer patients $6,000 to $8,000 per month. Given growing concerns about rising health-care costs, prices like these could limit adoption. At the same time, many patients may be wil ing to pay a high price for such potential y life-saving drugs. For advanced nanomaterials to deliver their ful potential through 2025 and beyond, reliable and far less expensive methods wil have to be developed for producing substances such as graphene, carbon nanotubes, and quantum dots in high volumes. Major chal enges persist in producing high-quality forms (long strands of nanotubes or large sheets of grapheme, for example) and effectively handling smal , delicate, chemical y reactive, and potential y toxic nanomaterials. Until these production chal enges can be overcome, the potential economic impact of advanced nanomaterials wil remain limited. Nanomaterials also raise a range of regulatory issues that wil need to be resolved before widespread adoption is possible. Some nanomaterials can have high toxicity and could cause environmental damage, and many remain untested. Regulations wil be needed to guide nanomaterial use not only in medicine, but also in other applications in which the material is not encapsulated. Products containing nanomaterials may require special end-of-life recycling procedures. 137 Matthew Herper, “Celgene’s Abraxane extends life by 1.8 months in advanced pancreatic cancer,” Forbes, January 22, 2013. 138 Nanotechnology for drug delivery: Global market for nanocarriers, Cientifica, February 7, 2012.
124 IMPLICATIONS Over the coming decade, advanced nanomaterials wil continue to attract considerable interest and R&D investment. These materials could also create major opportunities for health-care technology companies, pharmaceutical companies, and health-care providers. As the use of advanced nanomaterials becomes increasingly widespread, they have the potential to deliver enormous value to consumers, both in health care and eventual y across a wide array of products. However, policy makers wil need to address unanswered questions regarding the safety of nanomaterials. In industries such as semiconductors, consumer electronics, and chemicals, nanomaterials are already being explored, developed, and even used to produce some products. Nanomaterials are enabling these industries to move past the limitations that traditional materials impose on how fast a circuit can be or how efficient a chemical process can become. Across al of manufacturing, the cost/ benefit trade-offs of using nanomaterials for applications such as composites, paints, and coatings wil continue to shift as the technology evolves. For example, in the automotive industry, advanced nanocomposites and graphene-enhanced batteries could play a large role in enabling more cost-competitive electric vehicles. Business leaders should consider how these materials could ultimately be used to create revolutionary new products or make existing products better, and invest accordingly. As industries develop more products that use advanced nanomaterials—whether for electronic parts, composite materials, medicines, or other applications— companies that develop techniques for producing large quantities of high-quality nanomaterials like graphene and nanotubes could benefit greatly. Companies in many industries, particularly electronics and aerospace, should look for opportunities to invest in nanomaterials R&D, including through partnerships and sponsorships. Today, pure graphene producers are typical y smal er startup companies, such as Angstrom Materials, Graphenea, Graphene Square, Graphene Supermarket, Graphene Technologies, and XG Sciences. Carbon nanotubes also are produced by several major chemical companies, including Showa Denko, and Arkema. Consumers stand to benefit greatly from advanced nanomaterials. Besides offering potential breakthroughs in disease diagnosis and treatment over the coming decade, over the long term nanomaterials could also lead to new electronics products that are more powerful, more energy efficient, and more useful. Advanced nanocomposites using materials such as graphene and carbon nanotubes could eventual y be used to make many objects, including cars and airplanes, lighter and stronger. Nanomaterials might also help build a more sustainable future if they live up to their potential to create highly efficient batteries, solar cel s, and water purification systems. Policy makers wil need to consider both the costs and benefits of nanotechnology to their citizens, as wel as the economic implications of advanced nanomaterials. Large advanced economies have been funding nanotechnology research for two decades; in fact, in 2011 accumulated government investment over the previous decade totaled more than $67.5 bil ion. Including private investment, the total investment is estimated to be close to a
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 125 quarter of a tril ion dol ars.139 In 2011 China surpassed the United States as the largest funder of nanotechnology research. The objective of this research is to establish leadership and enable early adoption of nano-based products and processes by Chinese industry. Realizing the full, long-term potential of advanced nanomaterials wil require sustained funding and support. The US National Nanotechnology Initiative tracks and supports progress in key areas, but funding is vulnerable to congressional action and department budgets. Final y, serious studies need to be conducted to identify any environmental and health risks posed by nanomaterials. So far, research suggests that nanomaterials exhibit widely varying levels of toxicity depending on the type of materials and their configuration, with some materials appearing benign and others potential y highly toxic. As science and industry continue to pursue the economic and societal benefits of nanomaterials, policy makers and citizens should be informed about the potential environmental and health risks. 139 2011 report on global nanotechnology funding and impact, Cientifica, July 13, 2011.
126 11. Advanced oil and gas exploration and recovery After the first global oil shock during the 1970s, nations and energy companies around the world were forced to consider what a future with dwindling fossil fuel supplies might look like. One response was to look for new types of fossil fuel reserves and develop ways to reach them. Forty years later, these efforts are final y beginning to pay off. Horizontal dril ing and hydraulic fracturing, the technologies for reaching “unconventional” reserves such as the natural gas and “light tight” (LTO) oil trapped in rock formations (often shale) are now at the point of widespread adoption.140 These extraction techniques have the ability to unlock both newly discovered reserves and previously known deposits that could not be economical y extracted using conventional methods. North American energy companies, particularly those in the United States, are furthest along in exploration and development of unconventional reserves, and could maintain a sizable lead through 2025. Other countries are also believed to possess sizable reserves, but it could take many years to develop them.141 Global y, accessing these unconventional oil and gas resources could deliver significant economic impact by 2025. Shale gas production is already well under way, and LTO is becoming part of the North American oil supply. These developments wil clearly have an impact in the coming decade, which is why they are the focus of this analysis. However, they are not the only unconventional reserves; in fact, improved technologies could eventual y unleash yet another fossil fuel revolution. In this chapter, we examine the economic impact of advanced oil and gas exploration and recovery in five major regions with the potential to produce unconventional oil and gas in large quantities by 2025: North America, China, Argentina, Australia, and Europe. In these regions, we estimate that unconventional oil and gas could have a direct economic impact of $95 bil ion to $460 bil ion annual y by 2025.142 We estimate the bulk of this impact could come from North America due to the relative maturity of its industry. Moreover, most of that value could come from recovery of unconventional oil reserves due to greater incremental production over the coming decade (that is to say, oil reserves are starting from a smal er base than unconventional gas) and could have higher margins than shale gas. 140 Light tight oil is cal ed “light” because it is a variety of crude oil that is relatively lower in specific gravity; it is termed “tight” because it is found in deposits with low permeability. 141 World shale gas resources: An initial assessment of 14 regions outside the United States, US Energy Information Administration, April 2011. 142 Direct economic impact for this technology refers to the value added (GDP) to the economy by the oil and gas sector due to increased output from shale gas and LTO. Indirect impact refers to the value added to sectors of the economy that benefit from increased output in oil and gas, both upstream (oil-field equipment providers) and downstream (chemicals manufacturers). Induced impact refers to the value added to the economy due to increases in household incomes of people connected with the sector, such as employees of oil and gas companies and suppliers, as wel as their dependents and employees.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 127 In addition to providing direct economic impact, increased access to unconventional deposits could also create significant indirect and induced impact. Additional benefits could include lower input costs for downstream industries such as chemicals, which could raise output. However, recovering unconventional oil and gas could be constrained by political and regulatory actions spurred by environmental risks. As a result, realization of the ful potential economic impact may depend on choices made by societies and policy makers in the coming years. The implications of advanced oil and gas exploration and recovery are important and complex. Not only are there opportunities for energy players and suppliers, but there could also be substantial impact on downstream manufacturing industries and consumers. For some countries, these newly available resources promise greater self-reliance, which could alter their geopolitical postures. However, many citizens and leaders are concerned about the potential damage that hydraulic fracturing can have on local environments and ecosystems, leading some countries to ban the process entirely. There are also concerns that unconventional oil and gas could affect the development of renewable energy sources such as solar and wind by making fossil fuels once again cheaper by comparison. Over the coming decade, businesses, societies, and policy makers wil have to decide how best to benefit from new sources of fossil fuels while managing risks and concerns. If unconventional reserves can be safely and cost- effectively exploited, the pattern of growing energy scarcity may be reversed for many decades. DEFINITION Unconventional oil and gas reserves are defined as reserves that cannot be extracted by conventional dril ing methods. In these reserves, oil or gas is trapped in natural fractures in rock (often shale) or adsorbed by nearby organic material. Besides shale gas and LTO, unconventional fossil fuel deposits include coalbed methane, tight sandstone, and methane clathrates (also known as methane hydrates). Proven reserves of coalbed methane in Canada’s Alberta Province have been estimated to be equivalent to nearly 13 percent of the world’s shale gas reserves, and methane clathrate deposits are estimated to be many times larger than shale gas reserves.143 However, extraction of these reserves has thus far been difficult. The enormous methane clathrate deposits are located on the ocean floor, making them too expensive to recover in most cases.144 Development of coalbed methane has been set back by fal ing natural gas prices due to shale gas availability in North America. Therefore, this report focuses on only shale gas and LTO, which have the greatest potential for successful development through 2025, to estimate the economic impact of unconventional reserves. However, new technologies may lead to more rapid advances in the development of methane clathrates or coalbed methane, possibly ushering in the next energy “revolution.” 143 Potential of gas hydrates is great, but practical development is far off, US Energy Information Administration, November 2012; World shale gas resources, USEIA, April 2011; Alberta’s energy reserves 2007 and supply/demand outlook 2008–2017, Energy Resources Conservation Board, June 2008. 144 A notable promising development is the extraction of natural gas from methane clathrate deposits off the coast of Japan: “Japan extracts gas from methane hydrate in world first,” BBC News Business.
128 Today, the core technologies used to access unconventional oil and gas reserves are hydraulic fracturing and horizontal dril ing. When used together, these technologies are able to overcome many of the chal enges of extracting shale gas and LTO. Gas- and petroleum-rich shale rock is typical y located much deeper below the surface than conventional reserves (two to three miles). Because of shale’s low permeability, which prevents oil and gas from flowing from the rock, fracturing is required to release the pressure of overlying and surrounding rock. Fracturing involves pumping up to five mil ions of gal ons of fluid (usual y water- based with some additives) at high pressure into rock fractures to release gas or oil held in pores.145 Horizontal dril ing is the method by which the wel bore—the tube that carries the oil or gas up from the earth—is dril ed to the appropriate depth and then extended paral el to the surface up to a few kilometers. Horizontal dril ing al ows recovery of fuel in multiple stages along the length of the well bore, making it much more economical than dril ing repeatedly to a great depth (Exhibit 13).
Exhibit 13 Shale gas is located in pores and fractures 2 to 3 miles below the surface, requiring horizontal drilling and fracturing for extraction SOURCE: US Energy Information Administration; US Geological Survey In order to reach high levels of output from each shale basin and wel , large amounts of investment, experimentation, and data are required, often through a trial-and-error process of dril ing hundreds of trial wel s. Due to the varied characteristics of each basin, 500 to 1,500 wel s are typical y needed to ful y understand basin behavior. It can take up to $10 bil ion in capital investment and many years to scale up production of a basin, including building the infrastructure to transport the extracted oil or gas. 145 For more on the composition of fracturing fluids and additives, see Modern shale gas development in the United States, US Department of Energy, April 2009.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 129 POTENTIAL FOR ACCELERATION The potential for rapid development of shale gas and LTO is great given the world’s nearly insatiable appetite for fossil fuels to enable economic activity as populations grow and economies develop. As the massive populations of China and India grow richer, they can be expected to fol ow the rising pattern in per capita energy consumption that has been seen in other nations (Exhibit 14). Other drivers of demand for unconventional oil and gas are declining yields in some major fields and the desire by some nations to be more self-reliant for economic and political reasons.
Exhibit 14 As incomes rise, demand for resources increases; ENERGY EXAMPLE China and India may follow the same pattern Historic (1970–2008) 2030 projected Per capita energy consumption, 1970–2008, projected to 2030 for India and China Mil ion British thermal units per person 250 United States Germany United Kingdom South Korea 200 China Australia India Japan Indonesia France 150 Historical range for energy consumption 100 evolution 50 0 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000 Per capita GDP Real 2005 $PPP per person SOURCE: International Energy Agency; IHS Global Insight; McKinsey Global Institute analysis The technology for extracting unconventional oil and gas is advancing rapidly, pointing to the potential to significantly reduce costs and increase production. For example, it may be possible to double the productivity of fracturing by using microseismic data and wel log data in predictive fracture modeling. Such modeling techniques could cut the time it takes to understand basin behavior by half, enabling companies to scale up production faster. Replacing the diesel generators that power fracturing pumps with natural gas generators could reduce fuel costs for operating wel s and decrease nitric oxide and carbon dioxide emissions by up to 80 percent. Water reuse and treatment technologies could reduce freshwater needs by as much as 50 percent, saving up to $1 mil ion over the life of a wel . Fracturing with previously used and untreated water has helped reduce water treatment requirements and costs by more than 70 percent in the past few years. Longer term, use of non-water fluids such as vapor, refrigerated gas, or petroleum could increase productivity by as much as 20 percent per wel and make production easier in water-constrained areas (although the environmental effects of using alternative fracturing fluids are yet to be ful y studied).
130 Development of unconventional oil and gas fields is most advanced in the United States and Canada, but other nations are also beginning to develop their reserves. To get at China’s huge shale gas reserves, the Chinese government awarded exploration rights in January 2013 in 19 areas and has entered into an agreement with the US government to share technological know-how.146 To spur development in Argentina, the government has doubled the price of natural gas by raising its price cap and has invited foreign players to help explore unconventional reserves in Vaca Muerta, where conventional extraction has already spurred the creation of infrastructure. In Russia, which is believed to be one of the largest potential sources of unconventional oil and gas in the world, the energy industry has launched major exploration projects.147 The British government has al owed fracturing to resume in Lancashire (after it was stopped in 2012 fol owing tremors in nearby Blackpool), while South Africa and Romania have lifted their moratoria on exploration of shale fields.148 If these countries push aggressively to develop these resources, it is possible that the potential economic impact could be even greater than we are estimating here. POTENTIAL ECONOMIC IMPACT BY 2025 Over the long term, unconventional oil and gas could have significant economic impact on the global energy market. Based on current estimates of production for the five regions we studied—North America (the United States and Canada), China, Argentina, Australia, and Europe—we estimate potential direct economic impact of $95 billion to $460 billion a year by 2025 (Exhibit 15). The bulk of this potential direct economic impact could be in North America, where much of the incremental value would come from LTO production. Annual output of LTO could grow to five to seven times the current rate of production by 2025. Shale gas, by contrast, is already produced in large volumes in North America and therefore wil likely grow more moderately. Moreover, since oil prices are determined by international markets, increased local production has a smal er impact on prices and therefore on margins. Natural gas, meanwhile, is priced local y; in fact, the addition of shale gas to the supply has depressed prices in North America. Technical y recoverable reserves (known reserves) of shale gas are estimated to be 220 tril ion cubic meters (Tcm), equivalent to 1,350 bil ion barrels of oil, and unconventional oil reserves are estimated at 195 bil ion barrels.149 It should be noted that estimates of what is technical y recoverable have changed frequently as countries have explored their potential deposits. In North America, oil and gas deposits are wel documented and production in unconventional fields is under way; in the rest of the world, however, estimates of potential output are harder to predict and could change based on evolving national policies and changing understanding of the commercial viability of the basins being explored. It is very possible that estimates of reserves wil change significantly for many countries in coming years. 146 “Dozens line up for Chinese shale,” UP.com, January 22, 2013. 147 Anna Shiryaevskaya and Jake Rudnitsky, “Russia may hold 680 tril ion cubic meters of unconventional gas,” Bloomberg.com, December 24, 2012. 148 “UK gives backing for fracking—shale gas go ahead,” Materials World Magazine, February 2013. 149 One Tcm is equal to 35.31 Tcf (tril ion cubic feet) and 6.09 bil ion barrels of oil equivalent; for global reserves, see World shale gas resources, USEIA, April 2011.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 131
Exhibit 15 Sized applications of advanced oil and gas exploration and recovery could have direct economic impact of $95 billion to $460 billion per year in 2025 Potential economic Sized impact of sized Estimated potential regions and applications in 2025 Currently estimated incremental annual applications $ bil ion, annually reserves production in 2025 Assumed price in 2025 71 tril ion cubic meters 145 bil ion cubic $2–8 per mil ion British (Tcm) of reserves meters (Bcm) thermal unit (MMBtu); – 60 Tcm in nearly $70–280 mil ion United States per Bcm North America – 10– – 11 Tcm in Canada shale gas1 35
64 bil ion barrels of 5.4–9.0 mil ion $50–150 per barrel
reserves barrels per day – 57 bil ion barrels in
North America – 60– United States light tight oil 300 – 7 bil ion barrels in Canada More than 150 Tcm of 70–220 Bcm Regional pricing Rest of the world – 15– reserves (per MMBtu) shale gas 65 – 36 Tcm in China – China, Australia: $8–10 – 22 Tcm in Argentina – Argentina: $7–8 – Europe: $6–11 Rest of the world – 10– More than 130 bil ion 0.5–1.7 mil ion $50–150 per barrel light tight oil 60 barrels of reserves barrels per day – 24 bil ion barrels in Russia Other – 13 bil ion barrels in potential Argentina applications (not sized) Potential unsized applications include coalbed methane and methane clathrate Sum of sized potential 95– economic 460 impacts2 1 Potential economic impact estimated by calculating incremental gross output from 2025 production and prices, and converting into value added through GDP multiplier tables; currently estimated reserves are for information only. 2 Only direct value added—indirect and induced impact, as wel as downstream benefits, could nearly double the impact. NOTE: Potential economic impact not comprehensive;: includes potential impact of sized applications only. Numbers may not sum due to rounding. SOURCE: McKinsey Energy Insights; US Energy Information Administration; McKinsey Global Institute analysis Some of these estimates are highly sensitive to assumptions about the speed of development and future market prices. Our production estimates for how quickly nations wil gear up production could prove to be too conservative if these countries overcome regulatory, environmental, technological, and infrastructural (e.g., pipeline building) chal enges that typical y delay production. In addition, the price of oil in 2025 is dependent on many factors besides supply and demand, such as possible actions by current oil exporters to maintain price or the use of increased oil supply by new producers for geopolitical leverage. We have assumed a range of oil prices of $50 to $150 per barrel in 2025 for al countries (since it is a global y traded commodity) to show the sensitivity of the range of economic impact to this crucial assumption. 15
132 In addition to the direct economic impact created in the energy industry, advanced oil and gas exploration and recovery could yield indirect and induced economic impact that we estimate at between $85 bil ion and $420 bil ion annual y in 2025.150 For purposes of consistency with the other technologies in this report, we have focused on estimating the direct economic impact of extracting these reserves. In general, we have not estimated the impact on increasing output of other industries, though in the case of the United States, we have referenced a forthcoming McKinsey Global Institute report in laying out what it might be. North America (United States and Canada) Declining natural gas reserves during the 1970s prompted the United States government to fund research into extracting shale gas, leading to many advances in technology, including microseismic imaging. The government encouraged dril ing for shale gas through tax credits, research dissemination, and industry support. In 1991, it supported the first horizontal dril ing project, and in 1998, the first commercial shale fracture in the Barnett Shale basin in the US state of Texas. The first combination of hydraulic fracturing and horizontal dril ing fol owed in the Barnett basin in 2005. US development of unconventional oil and gas also benefits from the nation’s long history of oil exploration, more than a century’s worth of geological records, and a deep understanding of various technologies. Also, since any US landowner can sel mineral rights without government approval, it is relatively easy for developers to secure wel sites. Canada’s evolution in this regard has matched that of the United States, and it is currently the only other country producing shale gas or LTO in significant quantities. North American shale gas production has already reached 350 bil ion cubic meters (Bcm) annual y, supplying more than a quarter of its domestic natural gas production, and LTO production is currently about 1.5 mil ion barrels a day, or nearly 20 percent of total oil production. McKinsey’s proprietary Energy Insights model for unconventionals estimates that annual shale gas production could potential y reach 495 Bcm and that LTO could reach 6.9 mil ion to 10.5 mil ion barrels a day in 2025.151 To reflect the potential for great volatility in fuel prices, we have chosen a wide range for natural gas and oil prices to estimate impact. Therefore, having ranged shale gas prices from $2 to $8 per MMBtu and crude oil from $50 to $150 per barrel, we estimate direct economic impact on the North American economy from unconventional oil and gas at $70 bil ion to $335 bil ion per year by 2025.152 The indirect impact could be $30 bil ion to $150 bil ion, while the additional impact 150 For each country, we have used a combination of proprietary models, public sources, guidance from internal and external experts, and national planning targets to estimate the value added of incremental production of shale gas and oil by 2025. Using the same sources, we have ranged the price of these fuels in 2025 and used incremental production and price estimates to determine incremental gross output. We have then used standard GDP multiplier tables to convert gross output into incremental value added. 151 Energy Insights is part of McKinsey Solutions and provides distinctive analysis that enables energy players to make key decisions on strategy, investment, and performance. For more information on Energy Insights, visit http://solutions.mckinsey.com. 152 The unit of measurement MMBtu refers to mil ion British Thermal Units, traditional y the amount of energy required to heat one pound of water by one degree Fahrenheit, and often used for natural gas pricing; 1 MMBtu = 998.12 cubic feet.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 133 induced in the economy could be $45 bil ion to $220 bil ion. The total economic impact of increased shale gas and LTO production is therefore estimated to be between $145 bil ion and $705 bil ion in 2025. As mentioned earlier, an additional source of potential impact is the value added to the economy due to incremental output from energy-intensive industries such as petrochemicals. This is particularly relevant in the United States, as industry has benefited immensely from increased availability of cheap natural gas. According to research done for a forthcoming McKinsey Global Institute report on major sources of growth in the US economy, the petrochemicals industry is expected to add $60 bil ion to $80 bil ion of gross output in 2025 as a result of increased availability of cheap feedstock due to shale gas, and the manufacturing sector as a whole could add more than $100 bil ion in gross output, leading to bil ions in value added to the economy. In addition, other industries could be impacted; for example, in the long-haul transportation sector, some companies are now considering fueling trucks with natural gas.153 The rest of the world Many countries besides the United States and Canada are estimated to have substantial unconventional reserves and could realize significant economic benefits by developing them. However, most experts believe that by 2025 these nations wil reach only limited production. We have ranged our estimates of production and economic impact to reflect the differing views of experts, but it is possible that countries included in our estimates (such as China) or not included (such as Russia) could accelerate growth beyond our estimates. By 2025, the top producers outside North America are expected to be producing 70 Bcm to 220 Bcm of shale gas and 0.5 to 1.7 mil ion barrels per day of LTO. China could be among the biggest producers of shale gas in this group, while Argentina and Australia could be the biggest producers of LTO. Unlike crude oil, natural gas is not a freely traded commodity in international markets; its pricing is regional. Therefore, while we standardize on a global oil price estimate in 2025, we use regional estimates of shale gas prices to reflect conditions in different regions. Shale gas prices for China and Australia in 2025 are estimated at $8 to $10 per MMBtu, placing shale gas just below imported gas in price. Argentina’s government has mandated al new unconventional gas to be priced at $7.5 per MMBtu, while Europe’s natural gas pricing is more uncertain; we therefore use a wide range of $6 to $11 per MMBtu. The direct economic impact on these regions together is estimated to be $25 bil ion to $125 bil ion in 2025, with an additional $5 bil ion to $25 bil ion in indirect economic impact and nearly as much induced economic impact. 153 Diane Cardwel and Clifford Krauss, “Trucking industry is set to expand its use of natural gas,” The New York Times, April 22, 2013.
134 BARRIERS AND ENABLERS The extraction of unconventional deposits of oil and gas requires a range of enabling conditions and capabilities. If policy makers and societies are not aligned on the need for unconventionals, the ful impact of these resources wil not be realized. Large quantities of water (or advances in fracturing technologies that use other fluids) are required, as are access roads (often in remote areas) to move workers and equipment to dril ing sites. Several high-potential regions, such as China, Australia, and North Africa, suffer from water scarcity and could therefore find it difficult to al ocate water for fracturing. Australia and China also have limited transportation infrastructure in place to reach remote basins. The environmental risks of hydraulic fracturing, which include potential contamination of groundwater, air pol ution from equipment, greenhouse gas emissions from “fugitive” methane, and increased land use, are often cited as reasons for resistance to adoption of unconventionals. Basic horizontal dril ing and hydraulic fracturing technologies are not yet ful y developed in most regions outside North America. Recognizing this situation as a chal enge, the Chinese government recently signed a technology-sharing agreement with the United States.154 Access to capital is another potential barrier. Argentina has among the world’s largest unconventional reserves, but Argentinian companies could find it difficult to finance unconventional oil and gas development due to the country’s fiscal situation. National energy policies can work as either enablers or barriers. The United States government helped its industry take the lead by offering tax credits to companies that were wil ing to invest in shale exploration. Some Russian energy companies have been asking their government to consider similar tax credits for exploration.155 Meanwhile, many European countries impose economic barriers via the imposition of high costs and strict regulations. France and Bulgaria have imposed moratoria on shale gas exploration, although the United Kingdom and Romania have recently lifted such bans. IMPLICATIONS Businesses that participate in the energy value chain, especial y energy companies and oil-field services, could find enormous global opportunities in advanced oil and gas exploration and recovery. They could gain positions of strength by bringing the latest technological know-how to key players in these growing markets (often governments and state companies), and strike partnerships to explore unconventional basins together. Simultaneously, providers could consider investing in developing new techniques that could improve productivity or shorten learning curves for developing basins. For example, building expertise regarding water reuse could yield very high returns when production begins in earnest in water-scarce regions of the world such as parts of China or Africa. These players should also master big data analytics to improve their research. Shel , for example, is already col ecting up to a petabyte (one mil ion gigabytes) of geological data per wel using its advanced seismic monitoring sensors (codeveloped with Hewlett-Packard), and plans to use the sensors on 10,000 wel s. 154 “The US and China: Towards a Clean Energy Economy.” Whitehouse.gov., November 17, 2009. 155 Guy Chazan, “Red Lenin’ leads Russia’s oil revolution,” Financial Times, March 31, 2013.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 135 Other businesses in the unconventional oil and gas value chain can anticipate growth in demand for products and services and prepare accordingly. In addition to suppliers of fracturing equipment and dril ing rigs, new fields wil need cement for casings, steel for pipelines, water disposal systems, and geological consulting services. Meanwhile, downstream players such as petrochemical manufacturers can make decisions regarding locating new plant capacity or product planning in anticipation of gas being more abundantly available in some areas. Dow, BASF, and Methanex have already announced plans to set up new manufacturing capacity in the United States to take advantage of cheap natural gas prices. Businesses related to the renewable energy industry wil need to keep a close watch on the impact of unconventional oil and gas on solar and wind power development. Shale gas production has already helped drive natural gas prices in the United States down from over $10 per MMBtu in mid-2008 to less than $5 in 2013. The shift from coal to gas for electricity generation is reducing greenhouse emissions while simultaneously leaching urgency from the drive to develop renewable energy sources. However, it has been argued that greater deployment of natural gas instead of coal for power generation could create more “peaker” plants that can flexibly adjust their generation, replacing inflexible “baseload” plants (which are usual y coal-based). This could prepare grids for greater renewables capacity in the medium term by accommodating intermittent sources such as solar and wind power. Policy makers and citizens wil need to work together to weigh the benefits and risks of new oil and gas recovery technologies and determine whether they are, on aggregate, beneficial to society. Although the nature and extent of exposure is stil a topic of debate, exploration and recovery carry environmental risks, including potential contamination of groundwater and air and greenhouse gas emissions from fugitive methane. Technology for extracting unconventionals also requires huge amounts of resources—up to five mil ion gal ons of water in a single fracturing and hundreds of trips made by trucks to sites every day in the first few months. Many of these risks can be mitigated to a large extent through the use of new technologies and strong environmental policies. At the same time, development of shale plays can benefit areas where deposits exist, potential y creating jobs and economic development. If citizens, industry, and policy makers agree that exploiting unconventional reserves is a national economic and political priority, then governments could provide appropriate systematic support. Where reserves are not yet proven, governments could consider creating tax incentives for exploration and research. Land acquisition is a barrier in many nations, which governments can help to lower. In many European countries, al mineral wealth from approximately 1.5 meters (or five feet) below the surface and lower belongs to the state, reducing the incentive for land owners to permit access and forcing exploration companies to negotiate with the government—and the government to negotiate with landowners. Governments could consider changing these conditions to accelerate development. If resource-rich countries are able to support development of unconventional oil and gas by adopting the latest technologies, building necessary infrastructure, mitigating environmental risks, investing in research, and removing administrative obstacles, they could start their own shale revolutions, potential y realizing much larger economic impact than current estimates.
136 Access to energy sources has shaped the global geopolitical landscape for more than a century. The role of energy in defining foreign policy agendas could change as countries around the world begin to develop new local sources of energy. Some analysts have suggested that increased US production of unconventional gas and oil could affect its economic relations with energy exporters around the world and, therefore, could lead to a redefinition of US foreign policy priorities.156 Similarly, if Europe is able to develop large unconventional reserves or diversify its sources of supply with new LNG (liquid natural gas) exports from North America or other countries, its trade and economic relationship with Russia could be affected.157 No one can predict the geopolitical implications of such developments: there are simply too many variables in play. With that being said, governments, businesses, and citizens should anticipate that shifts in international energy flows wil have foreign policy implications. They should avoid any single “point prediction” and instead consider a range of possible scenarios when planning for the future. While acting on the possibilities and chal enges of the current situation, all stakeholders—including energy companies and suppliers, entrepreneurs, governments and policy makers—should be alert to the possibility that another revolution in unconventional energy production could be right around the corner. Just as the current potential has been made possible by incremental changes in technology over several years, the steps required to access the next big pool of value may already be under way. 156 Daniel Pipes et al., “The geopolitics of US energy independence,” International Economy, volume 26, issue 3, summer 2012. 157 “Unconventional gas in Europe: Frack to the future: Extracting Europe’s shale gas and oil will be a slow and difficult business,” The Economist, February 2, 2013.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 137 12. Renewable energy Renewable energy holds a simple but tantalizing promise: an endless source of power to drive the machinery of modern life without stripping resources from the earth; contributing to pol ution and climate change; or incurring the economic, social, and political tol associated with the competition for fossil fuels. This promise has been elusive because of the relatively high cost of renewable energy sources such as solar and wind compared with fossil fuels such as coal, oil, and gas. Despite some high-profile failures in the renewables sector and occasional y halting adoption, we see potential for rapidly accelerating growth in the next decade driven by both technological advances that could narrow the cost gap with fossil fuels and a growing desire to find energy sources that reduce human impact on the environment. While meeting future energy demands in a more sustainable way could involve many forms of renewable energy, solar and wind power could have particularly high potential over the coming decade. These sources of energy are final y beginning to be adopted at scale in advanced economies such as the United States and the European Union. Even more importantly, developing giants China and India have ambitious plans for renewable energy that could enable further rapid economic growth while mitigating growing concerns about pol ution. Although solar and wind energy could continue to be uncompetitive with fossil fuels on a pure cost basis in some regions through 2025, significant additional adoption is widely expected.158 This adoption could be driven by reduced costs, as wel as by continued government subsidies prompted by growing concerns over global climate change. In this chapter, we have estimated the potential economic impact of improving solar and wind technologies by comparing two scenarios—one that incorporates significant technological breakthroughs by 2025, and another in which technology is “frozen” at current levels. In comparing these two scenarios, we have included estimates of future government subsidies as a determinant of the levelized cost of electricity (LCOE), the metric that would be used when considering adoption.159 158 The International Energy Agency’s World energy outlook 2012 and Bloomberg’s Global renewable energy market outlook both estimate growth in renewables adoption from 4 to 5 percent to 11 to 15 percent. This is in line with the McKinsey Global Energy Perspective model estimates used in this chapter. 159 LCOE represents the overal cost of producing electricity, taking into account factors such as capital investment, operating and maintenance costs, capacity utilization, system efficiency, CO credits, exchange rates, and other factors. 2
138 We also consider multiple possible subsidy levels. The highest of these assumes a consensus among world governments to limit CO emissions to a specific 2 level in order to meet global climate change targets.160 Solar and wind power could represent 15 to 16 percent of global electricity generation in 2025, up from only 2 percent today. The incremental economic impact of this growth could be $165 bil ion to $275 bil ion annual y by 2025. Of this, $145 bil ion to $155 bil ion could be the direct value added to the world economy from this power, less the cost of subsidies.161 The remaining $20 bil ion to $120 bil ion per year reflects the possible value of the reduction in CO emissions. 2 Solar and wind power could generate enormous benefits for businesses that provide or consume energy, as wel as consumers and society, but this could still require strong government support, including continued subsidies. Increasing adoption of solar and wind power wil also be affected by patterns in fossil fuel prices, as wel as the actions of energy players, citizens, and policy makers. Greater demand for renewables could provide opportunities for technology providers and suppliers of ancil ary equipment. This demand could arise from concerns about environmental sustainability both by consumers and companies, including those adopting “triple bottom-line” approaches to governance (requiring social and environmental responsibility as wel as profits). Greater availability of renewables could create opportunities for companies to set up more environmental y responsible operations. For some businesses, solar or wind energy could make remote operations energy self-sufficient, expanding the range of possible locations. This would also depend to some degree on advances in energy storage (see Chapter 8). Utility companies could play a major role in the adoption of renewable energies. They may need to make some investments in storage capacity to accommodate intermittent flows of solar and wind-based power into their grids. However, distributed renewables—power bought from local, smal -scale operations or from commercial or residential users—could help defer investment in transmission and distribution infrastructure. Technological advances in renewable energies and in fossil fuel production are highly linked; progress in wind and solar power could reduce demand for fossil fuels, while advances in unconventional sources of oil and gas that expand supply and reduce prices (see Chapter 11, “Advanced oil and gas exploration and recovery”) could make renewables less competitive. However, it is also possible that greater use of gas to power electric plants could increase the share of peaker plants in the grid, making it more suitable for adoption of intermittent sources such as renewables in the medium and long term. An overarching question for the growth of wind and solar power is whether the environmental concerns of citizens wil be a sufficiently powerful force to motivate governments to continue 160 Most of the world’s major energy producers (except the United States and Canada) are signatories to the Kyoto Protocol and are legal y bound to reduce or limit their greenhouse gas emissions; the United States, Canada, and many other countries are signatories to the Copenhagen Accord, which includes nonbinding emission targets. 161 Direct economic impact here refers to the value added to the economy (GDP) due to increased output of electricity from solar and wind sources. The economy also benefits from indirect and induced impact, which we have not estimated here. Indirect impact refers to the value added to other sectors, such as manufacturers of transmission towers or gearboxes. Induced impact refers to the value added due to increases in the household incomes of people connected with the sector.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 139 to subsidize renewables (through environmental taxes on companies, for example) even if fossil fuel supplies rise and prices fal . DEFINITION Renewable energy is energy that is derived from a source that is continuously replenished, such as the sun, a river, wind, or the thermal power of the world’s oceans. While we have focused on solar photovoltaic (PV) technology and wind power in this chapter due to their potential to be disruptive over the next decade, other renewable sources of energy also have the potential to be transformative if technological development and adoption accelerate (see Box 13, “Other potential sources of renewable energy”). A source of renewable energy that we have not considered here is hydro-electric power, which holds a significant share of global power generation capacity but is general y not expected to see major technological advancements or increased adoption by 2025.162 Solar energy can be harnessed through many technologies, including concentrated solar and induced photosynthesis. We have sized the potential impact of the increased use of photovoltaics, the technology that accounts for the largest share of solar power production currently and which wil most likely do so in the future. PV panels are made of photosensitive materials such as crystal ine silicon. This technology converts sunlight into electric energy using the photoelectric effect.163 Solar panels can be used in smal arrays to power a single building or home, or deployed in massive solar “farms” that feed into the power grid. While the cost of solar PV cel s and the overal cost of solar power generation have dropped dramatical y in the past decade, solar power is stil not cost competitive with fossil fuels on a global basis, although in some regions it has achieved grid parity, or soon wil . The typical LCOE of conventional electric plants (coal and combined cycle gas) is around $50 per MWh, compared with nearly $150 per MWh for solar. Wind power has been harnessed by civilizations dating back to the ancient Egyptians. Modern wind power uses the same principles but attempts to maximize the effect with large blades and highly efficient turbines that turn more than 45 percent of the kinetic energy of wind into electricity. Wind farms can contain hundreds of turbines arrayed over hil tops, in rivers, or offshore in oceans. Offshore wind is stronger and more reliable than onshore, but it is more expensive to set up wind turbines in riverbeds or beneath the ocean. Western European countries are the leading users of wind power, fol owed by China. The LCOE for onshore wind, which includes CO credits, is now close to parity with coal and 2 oil, especial y in regions with large-scale wind power generation, but stil high on a global basis at $70 per MWh. The next waves of innovation in this technology may be focused on lowering the cost of wind turbines for offshore generation and designs that reduce the cost of nonturbine components, as wel as instal ation and maintenance costs. 162 The IEA’s World energy outlook 2012, Bloomberg’s Global renewable energy market outlook, and McKinsey’s Global Energy Perspective model all estimate the current share of hydro- electric power at 16 to 19 percent of global electricity generation and estimate it to remain within 15 to 18 percent in 2025. 163 The phenomenon in which electrical y charged particles are released from or within a material when it absorbs electromagnetic radiation, often defined as the ejection of electrons from a metal plate when light fal s on it.
140 Box 13. Other potential sources of renewable energy We have chosen to focus on solar photovoltaics and wind (both onshore and offshore) power for this report due to their potential for disruptive impact by 2025. But these are not the only important sources of renewable energy. Other technologies with significant potential include: Biofuels. Biofuels, which are made from organic materials such as corn, are currently not efficient sources of power and require a great deal of energy to harvest, transport, and process. The share of global energy supply from biomass is currently 2 percent and is projected to reach 4 percent in 2025. However, coupled with next-generation gene sequencing (see Chapter 6, “Next-generation genomics”), there is significant potential for increased power generation from biofuels. For example, research is under way to use photosynthesis from synthetical y sequenced cyanobacteria, a blue-green algae, to convert atmospheric CO into fuel. 2 Concentrated solar power (CSP). This technology uses giant lenses or mirrors to focus the sun’s energy on a smal area, converting light into heat energy to drive a steam turbine to produce electricity. CSP’s share of global electricity production is negligible, and investment in CSP has suffered as the price of photovoltaic cel s has fal en. Ocean thermal energy conversion. The world’s oceans are a huge untapped source of energy. Ocean thermal energy conversion (OTEC) uses the difference in temperature between deep and shal ow ocean water to generate electricity. There is immense energy in the world’s oceans; however, the technology to capture it is currently immature and produces only a very smal amount of power. Geothermal power. The temperature difference between the earth’s surface and its core can be used to drive a heat engine or steam turbine to produce electricity. Commercial y viable extraction is currently limited to only a few locations situated at tectonic plate boundaries (Iceland, for example) and is currently less than 1 percent of global power production. Next-generation nuclear power. Nuclear power already supplies nearly 15 percent of the world’s electricity. However, the technology faces significant social, political, economic, and environmental chal enges. Fol owing the 2011 Fukushima nuclear disaster in Japan, many countries have slowed, postponed, or canceled their nuclear programs. The question of nuclear waste storage is also a deterrent to adoption. However, next-generation nuclear technology taps into energy contained in current nuclear waste products to create a closed and sustainable system.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 141 Solar and wind are intermittent sources of energy, which complicates their use in utility grids. While solar energy is quite predictable, wind is much less reliable, causing difficulties in planning for maximum utilization and off-take. Both technologies are limited to certain locations: solar power generation requires abundant and unobstructed sunlight, while wind farms require large areas of land in locations where wind blows regularly. Current technologies for both require frequent—and often chal enging—maintenance. For example, solar panels need to be cleared of dust and debris regularly, especial y in deserts. Wind turbines often require maintenance of rotor components, which are hard to reach without specialized equipment due to the great height at which they are located. Solar energy peaks during the day and vanishes during the night, matching the typical demand cycle for power. Wind energy, on the other hand, is continuous throughout the day, though intermittent, and in some areas wind speeds are higher at night. It is therefore possible to combine solar and wind power to build a stronger combined portfolio by al owing the variability of one source to partial y offset that of the other. POTENTIAL FOR ACCELERATION The adoption of renewable sources of energy such as solar and wind power could be driven by two main factors: the need for greater amounts of energy to keep up with rising demand caused by economic growth and the need to mitigate environmental degradation and climate change. The world’s economic machinery requires a huge amount of energy, which could become more difficult to supply because of the nonrenewable nature of fossil fuels and the increasing costs of exploring new deposits. Even the unconventional oil and gas deposits that we discuss in Chapter 11, which could have significant impact on supplies in some nations, are exhaustible. Governments are increasingly aware of the need to ensure abundant energy supplies far into the future, which could potential y raise the demand for renewables. The environmental costs of reliance on fossil fuels are becoming more apparent every day. A recent study commissioned by China’s Ministry of Environmental Protection pegs the annual cost of damage to the ecosystem at $230 bil ion per year, or more than 3 percent of China’s GDP.164 In 2007 the United Nations Intergovernmental Panel on Climate Change (IPCC) projected that there wil be more cycles of extreme weather, that sea ice could shrink in both the Arctic and Antarctic, and that 20 to 30 percent of animal species studied could be at risk of extinction. Major economies such as the United States and China have agreed to target a maximum global temperature increase of two degrees Celsius by 2050 to limit these changes.165 As part of that effort, the United States has drawn plans to double its use of renewable energies by 2020, while the Chinese government has planned to meet at least 20 percent of the country’s energy demand with renewables by 2020.166 164 Edward Wong, “Cost of environmental damage in China growing rapidly amid industrialization, The New York Times, March 29, 2013. 165 The two-degree Celsius limit was agreed to by the United States and the BASIC nations (Brazil, South Africa, India, and China) as part of the Copenhagen Accord in 2009. 166 Evan Lehmann, Nathanael Massey, “Obama warns Congress to act on climate change, or he wil ,” Scientific American, February 2013; Jack Perkowski, “China leads the world In renewable energy investment,” Forbes, July 2012.
142 Technological advances that reduce the costs of renewable energy generation will be important enablers of adoption. During the past two decades, the efficiency of solar panels (the percentage of solar energy converted into electricity) has risen to 15 percent; in laboratory tests, panels have achieved as much as 44 percent efficiency.167 The cost of solar cel s has already dropped from nearly $8 per watt of capacity in 1990 to less than 80 cents.168 Wind technology is also progressing. Between 2000 and 2010, the average capacity of wind turbines in the United States doubled to 1.8 MW, as wind towers grew larger: hub height increased by a third to 260 feet and rotor diameter rose by two-thirds to more than 280 feet. The LCOE for wind power has fal en from nearly $80 per MW to $70 per MW, and wind generation costs now approach parity with coal and gas.169 Further advances in solar and wind power technologies are under way. Thin film cel s, which are made from compounds like cadmium tel uride, copper indium gal ium arsenide (CIGS), or amorphous silicon (A-Si), are being developed for PV use. These advances reduce the amount of material used in creating solar cel s and can be “printed” on flexible surfaces, potential y reducing cost and increasing ease of application. Researchers are also working with nanomaterials, including polymer films that are less than 100 nanometers thick that could replace silicon cel s and nanomaterial-based coatings that repel water and that prevent dust and debris from sticking to panels. Research is focused on turbine and blade design, such as the Japanese “Windlens,” which uses a curving ring around the perimeter of the blades to double or triple airflow through the blades. New vertical axis wind turbine designs place the main components—the generator, gearbox, and brake assembly—near the ground, making them more easily accessible for repair and maintenance. POTENTIAL ECONOMIC IMPACT BY 2025 As noted, we have estimated the potential economic impact of solar and wind power by comparing one scenario, which assumes significant technological breakthroughs by 2025, with a base scenario in which technology remains frozen at current levels; estimated impacts include the social value of CO emission 2 avoidance. In this way, the incremental impact of solar and wind technology, after netting the cost of government subsidies, could potential y be $165 bil ion to $275 bil ion a year in 2025 (Exhibit 16). The larger part of this impact could be in the form of direct value added to the economy from the power generated by the additional capacity created by solar and wind sources, as wel as the value added by the technological improvements that raise efficiency above current levels for these power sources. The remaining would be value to society from CO emission 2 avoidance. Only about 15 percent of the overal direct impact may come in North America, but more than 30 percent could come in China. (North America, on the other hand, currently stands to realize the most value from unconventional oil and gas.) 167 “Award-winning PV cel pushes efficiency higher,” National Renewable Energy Laboratory (NREL), www.nrel.gov/news/features/feature_detail.cfm/feature_id=2055. 168 Geoffrey Carr, “Sunny uplands,” The Economist, November 2012. 169 Mark Bolinger and Ryan Wiser, Understanding trends in wind turbine prices over the past decade, Lawrence Berkeley National Laboratory for US Department of Energy, October 2011.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 143
Exhibit 16 Sized applications of renewable energy could have economic impact of $165 billion to $275 billion per year in 2025 Potential economic Potential impact of sized productivity or Sized renewable renewables in 2025 Estimated scope Estimated potential value gains in energy sources1 $ bil ion, annually in 2025 reach in 2025 2025 1,330–1,570 TWh, 60–65% drop in or 5% of total the levelized cost Solar 105– 27,000–31,000 electricity generation of electricity photovoltaics 110 TWh global 1,100–1,300 TWh (LCOE) over electricity incremental base scenario consumption generation vs. base
8,500–9,500 TWh Cost 40– scenario renewables
impact Wind 45 generation (wind, 2,700–3,500 TWh, 25–30% drop in
solar, hydro) 10–11% of total LCOE over base $4.5–5.5 tril ion 500–550 TWh scenario annual generation Total 145– incremental – Onshore: cost of global cost impact 155 generation vs. base 13–15% electricity scenario – Offshore: 50% 2–degree Celsius maximum Solar 15– temperature rise 700–880 mil ion tons $20–100 per ton photovoltaics 90 target by 2050 of CO2 avoided Social 450 ppm global impact greenhouse gas (CO2 concentration limit avoided) 5– 280–300 mil ion tons $20–100 per ton Wind by 2050 30 of CO2 avoided Other Potential applications not sized include hydro, biomass, ocean potential thermal and wave energy, geothermal, next-generation nuclear, and applications concentrated solar power. (not sized) Sum of sized potential 165– economic 275 impacts2 1 Value calculated for a set of regions representing approximately 90% of the total market. 2 Only direct value added—indirect and induced impact not sized. NOTE: Potential economic impact not comprehensive; includes potential impact of sized applications only. Numbers may not sum due to rounding. SOURCE: McKinsey Global Energy Perspective; US Environmental Protection Agency; McKinsey Global Institute analysis Since subsidies are such an important factor in adoption of renewable energy sources, we tested two sets of assumptions about subsidies in the next decade. In one, subsidies increase moderately from current levels, with many countries offering no subsidies at al , even in 2025.170 The alternate scenario estimates the level of government subsidies that would induce enough use of renewables to meet climate change targets. These benefits are partial y offset by the cost of providing subsidies to energy producers. We estimate the incremental production in 2025 due to technology improvements (the difference between our technology- improvement scenario and our base scenario) and use national electricity prices and standard multiplier tables to translate impact into incremental value added. 170 Market prices for CO credits (or equivalent subsidies) are projected to rise in most markets 2 where they are already in place, acc 16 ording to the Int ernational Energy Agency’s World energy outlook 2012; the “low” end of our range of subsidies assumes regions with no subsidies (e.g. India, the Arab Gulf) wil continue to have no subsidies in 2025.
144 To estimate the benefit from avoided emissions, we have estimated a social cost of carbon.171 For our estimates, we have used a range of $20 to $100 per ton in 2025, based on a survey of estimates and roughly in line with third- party estimates.172 Current market prices are much lower (the European Union Emissions Trading Scheme price has dropped from $20 in 2008 to less than $5 in March 2013), and, we believe, do not reflect the cost of CO emissions to 2 society. We use the market price of CO as an input in the LCOE computations for 2 renewables but have not used it as the basis for our social cost estimate. Wind power capacity could be twice as much as solar power by 2025, but solar power could have greater economic impact. This is because of the greater potential for improvement in solar energy technology: in our technology- improvement scenario, the LCOE of solar power could fal by 50 percent or more by 2025, compared with the base scenario. Improvements in wind power technology could drive a reduction of up to 30 percent in the cost of wind power compared with the base scenario, with offshore wind, which is currently costly to instal and maintain, fol owing a steeper cost-reduction curve. Solar power The LCOE of solar power dropped from more than $400 per MWh in 2000 to $150 per MWh in 2010. We see potential for this rate of improvement to continue through 2025. The PV cel module and inverter, typical y 60 percent of the capital costs in these technologies, could fol ow a semiconductor-like improvement in price performance, while panel instal ation, usual y a fifth of the cost, can be made quicker and cheaper through GPS-guided power tools and robots. Overal , technology improvements could reduce solar LCOE by 65 percent by 2025. This drop could potential y accelerate adoption, raising total global solar capacity from 28 TWh in 2010 to between 1,330 and 1,570 TWh in 2025, or about 5 percent of global electricity production. Taking the incremental production over the base scenario (250 TWh) in 2025, the direct value added from solar power could be more than $100 bil ion a year, with an additional $15 bil ion to $90 bil ion in social impact through eliminated emissions. As described in the box below (see Box 14, “Distributed solar power”), solar power could also yield large benefits through distributed generation (use by consumers and businesses), especial y in cost savings from deferred investments in new transmission and distribution infrastructure. However, because of the intermittent nature of solar power, as wel as the need to accommodate bidirectionality (in which excess power generated during the day flows into the grid from end-points such as homes and other buildings), utilities may have to make some up-front investments to realize these medium- and long-term benefits. 171 The social cost of carbon is a metric that estimates “the monetized damages associated with an incremental increase in carbon emissions”; Michael Greenstone, Elizabeth Kopits, and Ann Wolverton, “Estimating the social cost of carbon for use in US federal rulemakings: A summary and interpretation,” Review of Environmental Economics and Policy, May 2011; Laurie T. Johnson and Chris Hope, “The social cost of carbon in US regulatory impact analyses: An introduction and critique,” Journal of Environmental Studies and Sciences, volume 2, number 3, August 2012; and Joanna M. Foster, “The social cost of carbon: How to do the math?” The New York Times, September 18, 2012. 172 Social cost of carbon for regulatory impact analysis under executive order 12866, US Environmental Protection Agency, February 2010; Wil iam D. Nordhaus, Estimates of the social cost of carbon: Background and results from the RICE-2011 model, National Bureau of Economic Research working paper number 17540, October 2011.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 145 Box 14. Distributed solar power In our estimates of impact, we have looked at only large-scale implementation of solar and wind power—wind and solar “farms” with hundreds or thousands of PV units or wind turbines that connect to power grids. However, distributed generation, principal y solar panels used to power individual households or supply part of a building’s energy requirements, may drive significant benefits in the coming years. Distributed generation enjoys a large share of overal solar power production in some countries, such as the United Kingdom and the Netherlands, where more than half of renewable generation is residential (mostly rooftop solar). The impact on national energy systems can be significant. A recent report by the California Solar Initiative estimated that 1 to 1.6 GW per year of solar power generated by consumers would supply the equivalent capacity of adding a new 500kV transmission line, estimated to cost nearly $1.8 bil ion in capital costs.1 Distributed generation could also provide other benefits, such as lower line losses due to shorter distances transmitted, productive use of unutilized real estate (rooftops), and environmental benefits. It could be particularly relevant for heavily congested areas where adding new infrastructure is impractical. Distributed generation can also make the grid more resilient, since it would continue to function when central infrastructure is out of commission. If renewable generation costs continue to fal and energy storage capabilities grow rapidly (see Chapter 8, “Energy storage”), we can imagine entire neighborhoods or factory complexes being served through distributed solar power. This could make remote housing and manufacturing plants more viable by reducing the transmission capacity required from the grid or even eliminating the need to access the grid altogether. Developing economies could benefit from electrifying “dark” vil ages or areas without incurring high costs in building infrastructure. 1 Joseph F. Wiedman, Erica M. Schroeder, and R. Thomas Beach, 12,000 MW of renewable distributed generation by 2020: Benefits, costs and policy implications, Interstate Renewable Energy Council, July 2012; California Solar Initiative 2009 impact evaluation, California Public Utilities Commission.
146 Wind power Given the relatively mature state of onshore wind power technology, we have focused on likely improvements in construction costs (potential y 20 percent) and operating expenses (potential y 25 percent). Offshore wind technology, which is much less evolved, could see a drop of more than 50 percent in capital expenditure requirements and operating expenses. For both onshore and offshore wind power, we base our estimates on improvements in the iron and steel materials used for turbine blades and advances in engineering design that enable longer and stronger blades, yielding greater output in slow wind, as well as reduced development costs achieved by increasing the scale of wind farms. For both onshore and offshore wind, we use the same social cost of CO as 2 for computation of the impact of solar power. The net effect of these factors, combined with an increase in the market value of CO credits, could potential y 2 reduce the LCOE of onshore wind by 15 percent and offshore wind by 50 percent compared with the base case. While the estimated drop for onshore wind is smal , it is significant given how close onshore wind already is to parity with fossil fuel plants. The direct value added for wind could be $40 bil ion to $45 bil ion per year, with an additional impact of $5 bil ion to $30 bil ion from CO emission avoidance. 2 Power generation through onshore as wel as offshore wind could grow significantly by 2025, from 330 TWh to 2,300–2,840 TWh for onshore and from six TWh to 400-650 TWh for offshore wind. Wind energy capacity could be distributed similarly to solar energy, with China, a major investor in wind power, accounting for nearly 30 percent of onshore wind production and advanced economies 40 percent. Together, wind generation could potential y account for 10 to 11 percent of the world’s electricity production by 2025. BARRIERS AND ENABLERS As noted, advances in other technologies are an important factor in the economics and adoption of renewable energy sources. While we are not assuming any significant advances in battery storage capabilities in our impact estimates, if storage technology does advance rapidly, it could make adoption of renewables, particularly distributed and off-grid solar and wind power, more economical. This could help bring electricity to remote or sparsely populated locations to which it is too difficult or expensive to extend transmission and distribution infrastructure. Battery storage also has the potential to make it more economical to generate power at scale in remote areas, for example through offshore wind or solar panel farms in deserts. Another factor for renewables, at least in places such as North America, could be the impact of unconventional oil and gas sources. An increase in the supply of gas and oil could drive down prices, as wel as reduce emissions, by substituting natural gas for coal, thereby reducing the economic and social rationale for adoption of renewables.173 At the same time, some studies conclude that greater availability of shale gas could lead to greater adoption of solar and wind power since the replacement of coal-fired “baseload” plants (that cannot easily adjust their output) with gas-fired “peaker” plants (which can easily cycle up and down) 173 Shale gas and light tight oil are fossil fuels found in source rock such as shale or which are adsorbed in nearby organic material, typical y three to five kilometers under the surface, and are extracted using horizontal dril ing and hydraulic fracturing technologies.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 147 could add the flexibility required for the grid to accommodate intermittent sources of energy like solar and wind power.174 As always with renewables, there is the question of subsidies. Despite technological advances and fal ing LCOEs, for most of the coming decade governments could need to subsidize solar and wind power to make them cost competitive with conventional fossil fuels such as coal and gas. While governments acknowledge the importance of continuing to push adoption in order to meet environmental goals, fiscal realities could make subsidies vulnerable to budget cuts. Germany, for example, reduced its solar feed-in tariffs (long-term purchase contracts based on the cost of generation, even when it is higher than market prices) by 15 percent in January 2012 and again by more than 20 percent in March 2012. Similar cuts in subsidies across Europe caused some companies to go out of business.175 Moreover, intergovernmental cooperation is required. Countries have sought to share the cost of combating climate change through mutual agreement, so that participating nations would shoulder the impact of subsidies and other costs of control ing global warming together. A lack of commitment from any of the major parties carries the danger of causing the rest to reduce their commitments. Conversely, coordinated action—for example, through an overhauled CO 2 pricing mechanism that reflects the true economic and social value of reducing emissions—could significantly accelerate adoption of renewables. IMPLICATIONS The prospect of renewables such as solar and wind taking on a much larger share of global energy generation has significant implications for energy players and related industries, for governments, and for citizens. Increasing demand for renewable power, especial y solar and wind, could generate opportunities for suppliers of renewable power across the world, as wel as suppliers and other businesses connected to the value chain. As adoption spreads, so wil the competition among global players. The Chinese solar and wind industries, which have made significant investments in the past decade, are wel positioned to grab a large share of the global market. But they could be chal enged by companies from other nations that are able to develop and commercialize next-generation designs and materials, as wel as improved operating and maintenance capabilities, more effectively. To accommodate the rising share of solar and wind power in grids, utilities could need to invest in infrastructure improvements to manage increased intermittency and bidirectionality (caused by feeds from distributed solar power) on their grids. These investments may seem costly, but utilities can better justify them if they take a portfolio view of their clean energy investments, which could include both solar and wind power (which are intermittent) and hydro-electric power (which is not). In addition, the two intermittent sources are complementary: in many places, wind power is stronger at night when solar is unavailable. Utilities could consider planning for greater renewables adoption by combining solar, wind, and hydro-electric power sources in a single portfolio and by investing in 174 Jason Channel , Timothy Lam, and Shahriar Pourreza, Shale & renewables: A symbiotic relationship, Citi Research, September 2012. 175 Mark Scott, “In Europe, green energy takes a hit from debt crisis,” The New York Times, November 2012.
148 advanced battery storage technologies to help accommodate renewables on the grid. This approach could not only make utilities cleaner, but also better able to meet peak demand. In many places, utilities can take advantage of solar power’s production pattern matching demand (during the day), driven by air-conditioning requirements. Distributed generation could even help defer infrastructure investments, reduce line losses, and add resiliency against centralized failures. Suppliers of fossil fuels may be adversely affected by the growth of renewables, which could curb demand for their products. They could also face government sanctions (carbon taxes, for example) that would add to their costs. Renewable energy sources could represent an opportunity for industries or companies that are heavy users of energy to undertake more environmental y sustainable operations to address the concerns of shareholders and other stakeholders. Consumers may not realize direct economic benefits from renewable energy since solar and wind power used by their utility providers may not cut their electric bil s. But they stand to gain in other ways. If renewables are adopted at the potential rates we describe and goals for reducing planetary warming can be achieved, the effects of climate change may be reduced and air and water quality could improve in many parts of the world. Reducing fossil fuel emissions could also have direct health effects, reducing the incidence of respiratory diseases and improving overal health through cleaner air, a need that is increasingly pressing in many rapidly developing nations. Governments wil have to weigh the benefits of adopting more solar and wind power against the costs of maintaining subsidies. To meet the global target of a less than two-degree rise in temperature by 2050, it is estimated that total carbon emissions cannot exceed 880 gigatons, of which 380 gigatons have already been used. At the current rate of emission growth, the 880-gigaton limit wil be reached in 2025.176 Policy makers may have to consider creating more aggressive policies related to greenhouse gas emissions and make more concessions toward developing international consensus on more environmental y sustainable uses of energy. 176 Malte Meinshausen et al., “Greenhouse-gas emission targets for limiting global warming to 2 °C,” Nature.com, April 30, 2009; James Leaton, “Unburnable carbon—Are the world’s financial markets carrying a carbon bubble?” The Carbon Tracker Initiative (www. carbontracker.org), 2009.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 149 Conclusion By the time the technologies that we describe are exerting their influence on the economy in 2025, it wil be too late for businesses, policy makers, and citizens to plan their responses. Nobody, especial y business leaders, can afford to be the last person using video cassettes in a DVD world. Business leaders, policy makers, and stakeholders should look ahead, identify the technologies that could affect them, and determine how to shape markets and policies in ways that will serve their interests. While the appropriate response wil vary by stakeholder and technology, we find that there are some useful guiding principles that can help define responses. BUSINESS LEADERS The technologies we have focused on in this report could fuel a decade of rapid innovation in products, services, business processes, and go-to-market strategies. Companies wil have new ways of developing and producing products, organizing their businesses, and reaching consumers and business-to-business customers. Business leaders wil need to determine when, how, and whether to take advantage of new technologies—and be prepared to move quickly when others use emerging technologies to mount chal enges. In the 21st century, al business leaders must understand technology. More than ever, leaders must develop their own wel -informed views of what developments like cloud computing could do for their enterprises and work to separate hype from reality. Leaders should think careful y about how specific technologies could drive economic impact and disruption in ways that could affect their businesses. Leaders should make sure to invest in their own technology knowledge; they don’t have to become programmers or compulsive Facebook posters, but they should keep abreast of technology trends and pay attention to what their most tech-savvy customers are doing and saying. A teenage customer halfway around the world may offer a better perspective on technology than a panel of experts in a conference room. Time is the enemy: the world is changing at Internet speed, and technology is continual y evolving. Strategies can quickly fal behind, so the rhythm of planning has to keep pace. When technologies have disruptive potential, the stakes are even higher and the range of strategic implications is wider. Management thinker Clayton Christensen warns companies against focusing too much on their largest, most established markets and related value propositions. In doing so, companies can miss the ways in which disruptive technologies can jump industry or market boundaries and change the rules of the game. The first MP3 files had inferior audio quality and were easily dismissed; they went on to make music CDs all but obsolete. When necessary, leaders must be prepared to disrupt their own businesses and make the investments to effect change: as the past two decades have shown, successful companies repeatedly reinvent themselves to keep up. This
150 wil require experimentation and investment. Early investment wil probably dilute the profitability of a company’s portfolio in the near term, but ultimately it is tomorrow’s sources of growth that ensure the enterprise’s future. In our experience, companies that real ocate resources early to capture trends often have higher returns and are more likely to survive long term. Failing to reinvent and focusing only on existing markets open the door for disruptors, particularly at the bottom end of the market. Everywhere, the democratization of technology is advancing, reducing barriers to entry and al owing entrepreneurs and other new competitors to disrupt established markets and industries. Cloud services make it easier for new companies with little capital to obtain operating infrastructure and access to markets that it has taken global companies decades to build. 3D printing goes a step farther; it not only opens up markets to competition from entrepreneurs, but it also has the potential to shift value directly to consumers as they learn how to make things that they used to buy. To compete in this environment, companies need deep sources of value or competitive advantage. The prospect of two bil ion new Internet users in developing economies promises both access to new consumers and the threat that those consumers wil go into business against you. To survive, companies wil need to learn the tricks of the Internet trade; multi- sided business models (such as online advertising or monetizing exhaust data) need not be reserved for Google and Facebook. Some of the biggest opportunities and chal enges for business leaders wil arise from new tools that could transform how work is done. With technologies like advanced robotics and automated knowledge work, companies could have unique opportunities to realize rapid improvements in productivity. These tools could redefine jobs as tasks are augmented by, or transferred to, machines, requiring new skil s for workforces. Knowledge workers are the foundation of future success—in al sectors of the economy; by 2025 some manufacturers could be hiring more designers and robotics experts than assemblers. Companies that use technology to make knowledge employees more productive wil gain large business model advantages and attract the best talent. More than ever, companies wil need to have the right people, along with the training systems to keep these workers’ skil s current. Adopting disruptive technologies entails risks, and managing these risks will be critical y important. Internal y, organizational effectiveness and cohesion could suffer as some jobs are transformed—or eliminated—by technology. By working with employees and redesigning jobs to focus on higher-value skil s— and by investing in workforce development—companies can minimize these risks. External risks include reputational risk, consumer resistance, as well as safety and regulatory issues. For example, new materials may have unforeseen health effects and may pose environmental risks. Autonomous vehicles might not deliver the potential impact we estimate unless the safety of driverless vehicles can be established, consumers accept the idea, and regulators come up with the necessary rules and standards to put these cars and trucks on the road. Business leaders need to strike a careful balance as they adopt new technologies; they must be thoughtful about risk, but they should also manage these risks without stifling potential.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 151 POLICY MAKERS Since the Industrial Revolution, governments have played an increasingly important role in bringing disruptive technologies to life. This ranges from the support for basic research that helped bring about the microelectronics and Internet revolutions to mobilizing coordinated efforts like the Human Genome Project. Equal y important, governments help set standards and facilitate the emergence of new markets, for example by setting the rules for the use of electromagnetic spectrum for things like mobile Internet devices or the Internet of Things. Governments also have the power to limit the adoption or progress of technology. The people who set policy have multiple responsibilities that are often in conflict when disruptive technologies emerge: the rising productivity that automation of knowledge work could enable could help drive productivity growth, but the potential impact on employment might create social and economic strain. At the same time, those who benefit from the established ways of doing things, imperiled by disruptive technologies, often find ways to influence policy. Given the scale of impact of the technologies we consider, the job of reconciling these conflicts and balancing the needs of today’s citizens with those of future generations wil place unprecedented demands on policy makers. Government leaders may need to update their approaches to facilitating adoption of new technologies while managing the attendant risks. Governments often provide initial funding and incentives for technology development and even act as early buyers to speed progress and adoption. In the past, government efforts often involved huge, decades-long projects. These large-scale programs could remain important going forward, but as technology grows ever more complex and competition between nations to be at the forefront of innovation intensifies, policy makers should consider models for smal er, more frequent experimentation in the development of specific technologies, while continuing to invest in basic research. At the same time, encouraging early adoption before technologies are completely understood raises some big challenges. For example, government support has been critical to the advance of nanomaterials and hydraulic fracturing and both carry potential health and environmental risks, which government also has the responsibility to address. In addition to creating incentives for the development and adoption of technologies, governments can play an important role in facilitating the creation of networks that can speed up innovation. By sponsoring col aborative efforts at a national or international level, governments can help bring world-class expertise to bear and foster relationships that extend beyond the research phase to help ensure successful commercialization. Standards-setting efforts could be very helpful to many of the disruptive technologies we describe, and governments can be influential in these processes. For example, the Internet of Things wil require a high level of inter-operability among different types of sensors and actuators across public and private networks. In addition, interfaces wil need to be secure to ensure that hackers and viruses do not interfere. This wil require international standard setting. Across technologies, governments wil need to cooperate on matters of international law, regarding intel ectual property, liability, and other issues that span borders.
152 The biggest chal enges for policy makers could involve the effects of technologies that have potential y large effects on employment. By 2025, technologies that raise productivity by automating jobs that are not practical to automate today could be on their way to widespread adoption. Historical y, when labor-saving technologies were introduced, new and higher value-adding jobs were created. This usual y happens over the long term. However, productivity without the innovation that leads to the creation of higher value-added jobs results in unemployment and economic problems, and some new technologies such as the automation of knowledge work could significantly raise the bar on the skil s that workers wil need to bring to bear in order to be competitive. Given the large numbers of jobs that could be affected by technologies such as advanced robotics and automated knowledge work, policy makers should consider the potential consequences of increasing divergence between the fates of highly skil ed workers and those with fewer skil s. The existing problem of creating a labor force that fits the demands of a high-tech economy wil only grow with time. Advanced economies are already facing a shortage of high-skill workers, particularly in technical fields. Secondary and tertiary curricula need to be aligned with those needs. Critical y, policy makers—as wel as employers—can no longer focus only on building the skil s of young people entering the labor force. They wil need to support the whole workforce, including through retraining. Policy makers also wil have opportunities to employ emerging technologies to address the chal enges of 2025. The mobile Internet and advances in the automation of knowledge work, for example, could make it possible to bring customized, interactive training to students and workers anywhere. Emerging technologies can also be used by government to deliver services more effectively and responsively. They can help societies address grand chal enges such as poverty and climate change. Final y, policy makers can take a step back and look at how they address technology issues. Technology continues to advance on many fronts and many developments bring new chal enges for society and government. Governments cannot afford to be passive or reactive. If policy makers wait until bioterrorists show what they can do with a low-cost gene-sequencing machine, it wil be too late. Policy makers should also have wel thought-out, structured methods for assessing technologies, to add rigor to their analyses. Also, they can look for better metrics to understand the value of emerging technologies. As we have discussed in this report, the value of technology often lies in its benefits to users—including in consumer surplus that is not captured by GDP measures. Measures of economic activity such as GDP are important and expedient, but better efforts can be made to measure aggregate societal and economic welfare comprehensively. Metrics can influence policy decisions, so policy makers should select them carefully.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 153 Bibliography Acemoglu, Daron, Simon Johnson, and James Robinson, “Institutions as a fundamental cause of long-run growth,” Handbook of Economic Growth, 2005. Atluri, Venkat, Umit Cakmak, Richard Lee, and Shekhar Varanasi, Making smartphones bril iant: Ten trends, McKinsey & Company Telecom, Media, and High Tech Extranet, June 2012. “Amazon apologizes for Netflix’s Christmas Eve streaming outage,” CNet, http:// news.cnet.com/8301-1023_3-57561454-93/amazon-apologizes-for-netflixs- christmas-eve-streaming-outage/. American Chemistry Council, “The NANO timeline,” Journal of Nanoparticle Research, volume 13, 2011. Annual energy outlook 2012, US Energy Information Administration, April 5–May 2, 2013. ASDReports, The aluminium market analysis, financials and forecasting 2012– 2017, May 2012. “Award-winning PV cel pushes efficiency higher,” National Renewable Energy Laboratory (NREL), www.nrel.gov/news/features/feature_detail.cfm/feature_ id=2055. Basu, Susanto, and John Fernald, “Aggregate productivity and aggregate technology,” Board of Governors of the Federal Reserve System: International Finance Discussion Papers, number 593, October 1997. Basu, Susanto, John Fernald, Miles Kimbal , “Are technology improvements contractionary?” The American Economic Review, volume 96, number 5, December 2006. BBC News Business, “Japan extracts gas from methane hydrate in world first,” www.bbc.co.uk/news/business-21752441. Bel , Gordon, “A Seymour Cray perspective,” Microsoft Research, November 1997. Bloomberg, Global renewable energy market outlook 2003, April 2013. Bolinger, Mark, and Ryan Wiser, Understanding trends in wind turbine prices over the past decade, Lawrence Berkeley National Laboratory for US Department of Energy, October 2011. Borm, Paul, and Wolfgang Kreyling, “Toxicological hazards of inhaled nanoparticles—Potential implications for drug delivery,” Journal of Nanoscience and Nanotechnology, volume 4, number 5, October 2004.
154 Bradley, Joseph, Joel Barbier, and Doug Handler, Embracing the Internet of Everything to capture your share of $14.4 tril ion, Cisco Systems, 2013. Broderick, Andrew, and David Lindeman, “Scaling telehealth programs: Lessons from early adopters,” Case Studies in Telehealth Adoption, The Commonwealth Fund, January 2013. Brynjolfsson, Erik, and Andrew McAfee, Race against the machine: How the digital revolution is accelerating innovation, driving productivity, and irreversibly transforming employment and the economy, 2011. Brynjolfsson, Erik, and Adam Saunders, “What GDP gets wrong (Why managers should care),” MIT Sloan Management Review, volume 51, number 1, Fall 2009. Bul is, Kevin, “How vehicle automation wil cut fuel consumption,” MIT Technology Review, October 24, 2011. Caughey, A. B., A. E. Washington, V. Gildengorin, and M. Kuppermann, “Assessment of demand for prenatal diagnostic testing using wil ingness to pay,” Obstetrics and Gynecology, volume 103, number 3, March 2004. California Public Utilities Commission, California Solar Initiative 2009 impact evaluation, June 2010. Cardwel , Diane, and Clifford Krauss, “Trucking industry is set to expand its use of natural gas,” The New York Times, April 22, 2013. Carr, Geoffrey, “Sunny uplands,” The Economist, November 2012. Channel , Jason, Timothy Lam, and Shahriar Pourreza, Shale & renewables: A symbiotic relationship, Citi Research, September 2012. Chazan, Guy, “‘Red Lenin’ leads Russia’s oil revolution,” Financial Times, March 31, 2013. Chen, Brian X., “The Samsung Galaxy S4 review roundup,” Bits blog, April 24, 2013, http://bits.blogs.nytimes.com/2013/04/24/samsung-galaxy-s4-review- roundup/. Chieh, Hang Chang, Yu Dan, and Chai Kah Hin, An exploratory study on understanding the technological dimension in disruptive innovation, The 5th International symposium on management of technology, 2007. Christensen, Clayton, “The ongoing process of building a theory of disruption,” Journal of Product Innovation Management, volume 23, 2006. Christensen, Clayton, and M.E. Raynor, The Innovator’s Solution: Creating and sustaining successful growth, 2003. Christensen, Clayton, Scott Anthony, and Erik Roth, Seeing what’s next: Using theories of innovation to predict industry change, 2004. Christensen, Clayton, Thomas Craig, and Stuart Hart, “The great disruption,” Foreign Affairs, March/April 2001. Cientifica, 2011 global nanotechnology funding and impact, July 2011.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 155 Cientifica, Nanotechnology for drug delivery: Global market for nanocarriers, February 2012. Cisco Systems, Cisco global cloud index: Forecast and methodology, 2011–2016, October 2012. Claude-Gaudil at, V., and B.C. Quelin, “Innovation, new market and governance choices of entry: The Internet brokerage market case,” Industry and Innovation, volume 13, 2006. Cohn, Jonathan, “The robot wil see you now,” The Atlantic, February 20, 2013. Cosgel, Metin, The political economy of law and economic development in Islamic history, University of Connecticut Department of Economics working paper number 2012–44, December 2012. Cowen, Tyler, The Great Stagnation: How America ate all the low-hanging fruit of modern history, got sick, and wil (eventual y) feel better, June 2011. Cutler, Kim-Mai, “Through dirt-cheap genetic testing, Counsyl is pioneering a new bioinformatics wave,” TechCrunch.com, April 23, 2013. David, Paul, “The dynamo and the computer: An historical perspective on the modern productivity paradox,” American Economic Review, volume 80, number 2, May 1990. David, Paul, “Zvi Griliches and the economics of technology diffusion: Linking innovation adoption, lagged investments, and productivity growth,” Stanford Institute for Economic Policy Research, March 20, 2011. Dil ow, Clay, “BMW augmented reality glasses help average Joes make repairs,” PopSci.com, September 3, 2009. Dittmar, J. E., “Information technology and economic change: The impact of the printing press,” The Quarterly Journal of Economics, volume 126, number 3, 2011. Drummond, Katie, “Clothes wil sew themselves in Darpa’s sweat-free sweatshops,” Wired, June 8, 2012. Dumaine, Brian, “The driverless revolution rol s on,” Fortune, November 12, 2012. Dundas, Jitesh, “Machine learning helps software development,” Software Magazine, June 2012. El-Kady, Maher, and Richard Kaner, “Scalable fabrication of high-power graphene micro-supercapacitors for flexible and on-chip energy storage,” Nature Communications, volume 4, February 2013. Energy Resources Conservation Board, Alberta’s energy reserves 2007 and supply/demand outlook 2008–2017, 2008. English, Andrew, “New car tech: 2014 Mercedes-Benz S-class,” Road & Track, November 2012. Enterprise surveys: What businesses experience, World Bank Group.
156 Eurostat, The contribution of renewable energy up to 12.4% of energy consumption in the EU27 in 2010, press release, June 18, 2010. Evans, Dave, The Internet of Things: How the next evolution of the Internet is changing everything, Cisco, April 2011. Ervin, Matthew H., Carbon nanotube and graphene-based supercapacitors: Rationale, status, and prospects, Army Research Laboratory, August 2010. Foster, Joanna M., “The social cost of carbon: How to do the math?” The New York Times, September 18, 2012. Gantz, John, and David Reinsel, The digital universe in 2020: Big data, bigger digital shadows, and biggest growth in the Far East, International Data Corporation (IDC) analyst report, December 2012. Glanz, James, “Power, pol ution and the Internet,” The New York Times, September 22, 2012. Google, The impact of clean energy innovation: Examining the impact of clean energy innovation on the United States energy system and economy, July 2011. “Google announces Provo, Utah as the third Google Fiber city and acquires the local fiber provider,” TechCrunch, http://techcrunch.com/2013/04/17/google- announces-provo-utah-as-the-third-google-fiber-city-acquires-the-local-fiber- provider/. Gordon, Robert, “Is US economic growth over? Faltering innovation confronts the six headwinds,” National Bureau of Economic Research, August 2012. Govindarajan, Vijay, and Praveen Kopal e, “Disruptiveness of innovations: Measurement and an assessment of reliability and validity,” Strategic Management Journal, volume 27, 2006. Greenstein, Shane, “Measuring consumer surplus online,” The Economist, March 2013. Greenstone, Michael, Elizabeth Kopits, and Ann Wolverton, “Estimating the social cost of carbon for use in US federal rulemakings: A summary and interpretation,” Review of Environmental Economics and Policy, May 2011. Grocott, Michael, Measuring morbidity fol owing major surgery, 2010. Hanlon, Michael, “World’s first GM babies born,” Daily Mail, April 27, 2013. Hart, Stuart, and Clayton Christensen, “The great leap: Driving Innovation from the base of the pyramid,” MIT Sloan Management Review, October 15, 2002. Hendersen, R.M., and K.B. Clark, “Architectural innovation: The reconfiguration of existing product technologies and the failure of established firms,” Administrative Science Quarterly, volume 35, 1990. Hensel B., “Pil shaped camera goes where doctors cannot,” video, RedOrbit website, January 2013.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 157 Hensley, Russel , John Newman, and Matt Rogers, “Battery technology charges ahead,” The McKinsey Quarterly, July 2012. Herper, Matthew, “Celgene’s Abraxane extends life by 1.8 months in advanced pancreatic cancer,” Forbes, January 22, 2013 Heussner, Ki Mae, Recycle or else? Big Brother to pick through trash, ABC News, August 26, 2010. Hutton, David, Older people in emergencies: Considerations for action and policy development, World Health Organization, 2008. Independent Evaluation Group (IEG) and World Bank, The welfare impact of rural electrification: A reassessment of the costs and benefits, 2008. International Data Corporation, Worldwide quarterly mobile phone forecast, March 2012. International Energy Agency, World energy outlook 2012, November 2012. International Energy Agency, World energy outlook 2011, November 2011. International Energy Agency/Energy Technology Systems Analysis Program and International Renewable Energy Agency, Electricity storage, technology policy brief E-18, April 2012. International Federation of Robotics, 68 robots perform farmer’s work, case study of Fanuc Robotics Europe S.A., September 2012. International Federation of Robotics, Positive impact of industrial robots on employment, February 21, 2011. International Federation of Robotics, World robotics 2012, August 30, 2012. John, Rijo, and Hana Ross, The global economic cost of cancer, American Cancer Society and Livestrong, 2010. Johnson, Laurie, and Chris Hope, “The social cost of carbon in US regulatory impact analyses: An introduction and critique,” Journal of Environmental Studies and Sciences, volume 2, number 3, 2012. Kostoff, Ronald, Robert Boylan, and Gene Simons, “Disruptive technology roadmaps,” Technological Forecasting and Social Change, volume 71, 2004. Krantz-Kent, Rachel, “Measuring time spent in unpaid household work: Results from the American Time Use Survey,” Monthly Labor Review, volume 132, number 7, July 2009. Kumparak, Greg, “Google announces Provo, Utah as the third Google Fiber city and acquires the local fiber provider,” TechCrunch.com, April 17, 2013. La Vigne, Nancy G., Samantha S. Lowry, Joshua A. Markman, and Al ison M. Dwyer, Evaluating the use of public surveil ance cameras for crime control and prevention, Urban Institute and Community Oriented Policing Services, September 2011.
158 Ladd, Thaddeus D., Fedor Jelezko, Raymond Laflamme, Yasunobu Nakamura, Christopher Monroe, and Jeremy L. O’Brien, “Quantum computers,” Nature, volume 464, number 7285, March 2010. Lanier, Stephanie, “Hidden features and sensors in the Samsung Galaxy S4,” Android Guy blog, April 10, 2013, www.androidguys.com/2013/04/10/hidden- features-and-sensors-in-the-samsung-galaxy-s4/. Lardinois, Frederic, “Google’s self-driving cars complete 300K miles without accident, deemed ready for commuting,” TechCrunch, August 7, 2012. Leaton, James, “Unburnable carbon—Are the world’s financial markets carrying a carbon bubble?” The Carbon Tracker Initiative (www.carbontracker.org), 2009. Lehmann, Evan, and Nathanael Massey, “Obama warns Congress to act on climate change, or he wil ,” Scientific American, February 2013. Liefer, Richard, et al., Radical Innovation, 2000. Leong. Jasmine, “New method for continuous production of carbon nanotubes by USM researchers,” Scientific Malaysian, April 23, 2012. Leunig, Tim, “Social Savings,” Journal of Economic Surveys, volume 24, issue 5, December 2010. Lockheed Martin, “Lockheed Martin achieves patent for perforene filtration solution, moves closer to affordable water desalination,” press release, March 18, 2013. Lowensohn, Josh, “Eyes-on: Samsung’s Youm flexible-display tech at CES 2013, CNET blog, January 9, 2013, http://ces.cnet.com/8301-34435_1-57563058/eyes- on-samsungs-youm-flexible-display-tech-at-ces-2013/. MacKenzie, Angus, “The future is here: Piloted driving in Audi’s autonomous A6,” MotorTrend, January 14, 2013. Maddison, Angus, “Causal influences on productivity performance 1820–1922: A global perspective,” Journal of Productivity Analysis, November 1997. Meinshausen, Malte, Nicolai Meinshausen, Wil iam Hare, Sarah C. B. Raper, Katja Frieler, Reto Knutti, David J. Frame, and Myles R. Al en, “Greenhouse-gas emission targets for limiting global warming to 2 °C,” Nature.com, April 30, 2009. Markoff, John, “Armies of expensive lawyers, replaced by cheaper software,” The New York Times, March 4, 2011. Markoff, John, “Scientists see promise in deep-learning programs,” The New York Times, November 23, 2012. Markoff, John, “Skil ed work, without the worker,” The New York Times, August 18, 2012. McKinsey & Company, Online and upcoming: The Internet’s impact on aspiring countries, January 2012.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 159 McClain, Dylan Loeb, “First came the machine that defeated a chess champion,” The New York Times, February 16, 2011. McKinsey & Company, Personalized medicine: The path forward, March 2013. McKinsey & Company for IAB (Internet Advertising Board) Europe, Consumers driving the digital uptake: The economic value of online advertising-based services for consumers, September 2010. McKinsey Global Institute, Growth and renewal in the United States: retooling America’s growth engine, February 2011. McKinsey Global Institute, Internet matters: The Net’s sweeping impact on growth, jobs, and prosperity, May 2011. McKinsey Global Institute, The social economy: Unlocking value and productivity through social technologies, July 2012. McKinsey Global Institute, The world at work: Jobs, pay, and skil s for 3.5 bil ion people, June 2012. McWil iams, Andrew, Graphene: Technologies, applications and markets, BCC Research, July 2012. Measurement Incorporated, “MI’s automated essay scoring system takes first prize in national competition,” press release, October 4, 2012. Mel , Peter, and Timothy Grace, The NIST definition of cloud computing, National Institute of Standards and Technology, Special Publication number 800–145, September 2011. Meter, Ken, Hoosier farming? Emergent food systems in Indiana, Indiana State Department of Health, January 2010. Military.com, Shortage of truck drivers and federal funding, March 21, 2013. Mil er, Michael J., “A CES surprise: Mobile processor announcements,” Forward Thinking blog, January 17, 2013, http://forwardthinking.pcmag.com/none/307104- a-ces-surprise-mobile-processor-announcements. Mims, Christopher, “AI that picks stocks better than the pros,” MIT Technology Review, June 10, 2010. Mohdahl, Mary, Tablets set to change medical practice, Quantia MD report, June 2011. Mokyr, Joel, “Long-term economic growth and the history of technology,” Handbook of Economic Growth, 2005. Musil, Stephen, “Amazon apologizes for Netflix’s Christmas Eve streaming outage,” CNet.com, December 31, 2012. National Aeronautics and Space Administration, A growing intel igence around earth, Science News, October 2006.
160 National Aeronautics and Space Administration, Nanotechnology roadmap, April 2012. National Aeronautics and Space Administration, Robotics, tele-robotics and autonomous systems roadmap, April 2012. Nieto, Marcus, Public video surveil ance: Is it an effective crime prevention tool? California Research Bureau, 1997. Nordhaus, Wil iam D. Estimates of the social cost of carbon: background and results from the RICE-2011 model, National Bureau of Economic Research working paper number 17540, October 2011. Norotte, Cyril e, F. S. Marga, L. E. Niklason, and G. Forgacs, “Scaffold-free vascular tissue engineering using bioprinting,” Biomaterials, volume 30, number 30, October 2009. Oliver, John, Global markets and technologies for carbon nanotubes, BCC Research, July 2012. Paap, J., and R. Katz, “Anticipating disruptive innovation,” Research Technology Management, volume 27, 2004. Perkowski, Jack, “China leads the world in renewable energy investment,” Forbes, July 2012. Peterson, Scott, Jay Whitacre, and Jay Apt, “The economics of using plug-in hybrid electric vehicle battery packs for grid storage,” Journal of Power Sources, volume 195, number 8, April 2010. Phil ips, Lisa E., “Trends in Consumers’ Time Spent with Media,” eMarketer.com, December 28, 2010. Rothaermel, F.T., “Incumbent’s advantage through exploiting complementary assets via inter-firm cooperation,” Strategic Management Journal, volume 22, 2001. Rubin, Jared, “The printing press, reformation, and legitimization,” Stanford, Stanford University. Pipes, Daniel, et al., “The geopolitics of US energy independence,” International Economy, volume 26, issue 3, summer 2012. Pol ack, Andrew, “Engineered fish moves a step closer to approval,” The New York Times, December 21, 2012. Rastler, D., Electricity energy storage technology options: A white paper primer on applications, costs, and benefits, Electric Power Research Institute white paper, December 2010. Rio Tinto Iron Ore press release: www.riotintoironore.com. Roco, Mihail C., “The long view of nanotechnology development: The National Nanotechnology Initiative at 10 years,” Journal of Nanoparticle Research, volume 13, number 2, February 2011.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 161 Roco, Mihail C., and Wil iam Sims Bainbridge, Societal implications of nanoscience and nanotechnology, National Science Foundation, March 2001. “Researchers testing frugal autonomous car system, aim for $150 price tag,” video, Engadget. “Russia may hold 680 tril ion cubic meters of unconventional gas,” Bloomberg, www.bloomberg.com/news/2012-12-24/russia-may-hold-680-tril ion-cubic- meters-of-unconventional-gas.html. Schmidt, G., and C. Druehl, “When is a disruptive innovation disruptive?” Journal of Product Innovation Management, volume 23, 2006. Scott, Mark, “In Europe, green energy takes a hit from debt crisis,” The New York Times, November 13, 2012. Scudel ari, Megan, “Missing touch,” The Scientist, September 1, 2012. SDG&E Sunrise Powerlink Web Resources, http://regarchive.sdge.com/ sunrisepowerlink. “Shining a light,” The Economist, March 8, 2007. Sklar, Kathryn Kish, Florence Kel ey and the nation’s work: The rise of women’s political culture, 1830–1900, Yale University Press, 1995. SmartAction, “Roadside assistance customers benefit from smart support during peak and after hours,” case study of Canadian Automobile Association Saskatchewan, June 2012. Sood, Ashish, and Gerard Tel is, “Demystifying disruption: A new model for understanding and predicting disruptive technologies,” Marketing Science, 2010. Tasoglu, Savas, and Utkan Demirci, “Bioprinting for stem cel research,” Trends in Biotechnology, volume 31, number 1, January 2013. “The great innovation debate,” The Economist, January 12, 2013. Thomas, R., and M. Barnes, “Life expectancy for people with disabilities,” NeuroRehabilitation, May 2007. “Through dirt-cheap genetic testing, Counsyl is pioneering a new bioinformatics wave,” TechCrunch, http://techcrunch.com/2013/04/23/counsyl/. Tohani, Malavika, Global lithium-ion battery market: Growth trends and application analysis, Frost & Sul ivan, February 2013. “Trends in consumers’ time spent with media,” eMarketer, www.emarketer.com/ Article/Trends-Consumers-Time-Spent-with-Media/1008138. Trimble, Stephen, “Lockheed Martin reveals F-35 to feature nanocomposite structures,” Flight International, May 26, 2011. “Unconventional gas in Europe: Frack to the future: Extracting Europe’s shale gas and oil wil be a slow and difficult business,” The Economist, February 2, 2013.
162 “UK gives backing for fracking: What’s next?” Mil icent Media, http:// mil icentmedia.com/2013/04/10/uk-gives-backing-for-fracking-whats-next/. United Nations Framework Convention on Climate Change, Kyoto Protocol, December 1997. UPI, “Dozens line up for Chinese shale,” www.upi.com/Business_News/Energy- Resources/2013/01/22/Dozens-line-up-for-Chinese-shale/UPI-94351358857377/ US Department of Energy, Modern shale gas development in the United States, April 2009. US Department of Energy, SGIG accelerates grid modernization in Minnesota, Smartgrid.gov. US Energy Information Administration, Potential of gas hydrates is great, but practical development is far off, November 2012. US Energy Information Administration, World shale gas resources: An initial assessment of 14 regions outside the United States, April 2011. US Environmental Protection Agency, Social cost of carbon for regulatory impact analysis under executive order 12866, February 2010. V., John, “Apple iPhone 4S benchmark tests,” phonearena.com, October 20, 2011. Varian, Hal, “The value of the Internet now and in the future,” The Economist, March 10, 2013. Vienna University of Technology, “3D-printer with nano-precision,” press release, March 12, 2012. Von Holst, Hans, ed., Transportation, traffic safety and health: The new mobility, Springer, 1997. Wal , Richard, Jerry Bennett, Greg Eis, Kevin Lichy, and Elizabeth Owings, Creating a low-cost autonomous vehicle, presented at IEEE, Industrial Electronics Society annual conference in Sevil a, Spain, November 5–8, 2002. Weintraub, Karen, “Gold particles could deliver cancer drugs,” Boston Globe, December 24, 2012. Weiser, Thomas G., Scott E. Regenbogen, Katherine D. Thompson, Alex B. Haynes, Stuart R. Lipsitz, Wil iam R. Berry, and Atul A. Gawande, “An estimation of the global volume of Surgery: A modeling strategy based on available data,” Lancet, volume 372, number 9633, July 12, 2008. White, Corey D., and K. Max Zhang, “Using vehicle-to-grid technology for frequency regulation and peak-load reduction,” Journal of Power Sources, volume 196, number 8, April 2011. Wiedman, Joseph F., Erica M. Schroeder, and R. Thomas Beach, 12,000 MW of renewable distributed generation by 2020: Benefits, costs and policy implications, Interstate Renewable Energy Council and Crossborder Energy, July 2012.
McKinsey Global Institute Disruptive technologies: Advances that will transform life, business, and the global economy 163 Wohlers Associates, Wohlers report 2012: Additive manufacturing and 3D printing state of the industry, 2012. Wong, Edward, “Cost of environmental damage in China growing rapidly amid industrialization,” The New York Times, March 29, 2013. Yu-Ming Lin, Alberto Valdes-Garcia, Shu-Jen Han, Damon B. Farmer, Inanc Meric, Yanning Sun, Yanqing Wu, Christos Dimitrakopoulos, Alfred Gril , Phaedon Avouris, and Keith A. Jenkins, “Wafer-scale graphene integrated circuit,” Science, volume 332, number 6035, June 2011. Zebra Technologies, Building value from visibility: 2012 enterprise Internet of Things adoption outlook, Forrester Consulting, June 2012.
Related McKinsey Global Institute research Urban world tablet app for iPad and Android (May 2013) An unprecedented wave of urbanization is driving the most significant economic transformation in history, as the center of economic gravity shifts decisively east. MGI’s new tablet app, now available for Android as wel as iPad, offers an intuitive sense of this new urban world, showcasing GDP, population, and income levels for over 2,600 cities worldwide in 2010 and 2025. The app is available from the Apple App Store and Google Play. The social economy: Unlocking value and productivity through social technologies (July 2012) MGI analyzes the potential impact of social technologies. Based on in-depth analyses of how social technology can be used in five economic sectors (four commercial sectors, plus the social sector), the report identifies a series of value levers that can be applied across the value chain and within and across enterprises. Online and upcoming: The Internet’s impact on aspiring countries (January 2012) This report explains how the Internet today connects about two bil ion people worldwide. Half of these are in the “aspiring” world—countries as varied as Algeria, South Africa, China, Iran, and Mexico that are climbing the developmental ladder quickly, with diverse populations and inarguable economic potentialities. It examines the impact of the Internet in populous and fast-growing aspiring countries, where it offers even greater potential than in the developed world. Big data: The next frontier for innovation, competition, and productivity (May 2011) Big data wil become a key basis of competition, underpinning new waves of productivity growth, innovation, and consumer surplus—as long as the right policies and enablers are in place. Internet matters: The Net’s sweeping impact on growth, jobs, and prosperity (May 2011) The Internet is a vast mosaic of economic activity, ranging from mil ions of daily online transactions and communications to smartphone downloads of TV shows. But little is known about how the Web in its entirety contributes to global growth, productivity, and employment. McKinsey research into the Internet economies of the G-8 nations as wel as Brazil, China, India, South Korea, and Sweden finds that the Web accounts for a significant and growing portion of global GDP. www.mckinsey.com/mgi E-book versions of selected MGI reports are available at MGI’s website, Amazon’s Kindle bookstore, and Apple’s iBookstore. Download and listen to MGI podcasts on iTunes or at www.mckinsey.com/mgi/publications/multimedia/