The Firm, The University and The Regulator

Geographic Instigators and Implications of AV R&D

The following research project was completed over the academic year of 2016/2017 at the University of Toronto as part of a year-long internship with Urban Studies at the Martin Prosperity Institute. The opinions expressed here represent the work of our intern, under the supervision of Richard Florida, Shauna Brail, and Karen M. King, and do not necessarily reflect those of the Martin Prosperity Institute.

Download this Report (PDF)

Cities are inseparably linked to the transportation networks that run through them. Our urban realms and global economic structures have already been revolutionized twice following the introduction of rail and then the automobile. Both had significant impacts on the ordering of space, land, and capital, allowing for connected transportation routes and the possibility for cities to sprawl out of their cores.

Today, we are on the cusp of the next transportation revolution; autonomous vehicles (AVs) will set the stage for the future of international industry competition and cooperation, as well as the reordering of our social and physical spaces. Although automated options like cruise control, parking assistance, and proximity sensors are not new, AVs, connected to each other through an algorithmic grid and operated without human operation or intervention, are set to revolutionize the way our cities function.1 The AV revolution is set to become reality over the next few decades, with consumers beginning to widely adopt AVs around 2030; by 2050, AVs are expected to become the primary form of transportation.2

The AV industry is still in its infancy and current research and development (R&D) initiatives are central to a number of critical hurdles the industry still needs to pass over, including safety and public education around the technology. While the AV R&D arena is full of heavyweight incumbent automotive and technological firms fighting for trailblazing status, behind the scenes, university talent and academic-industry partnerships are playing a decisive role. This wave of AV R&D will have tremendous transformative results in the academic and industrial institutions that possess the locational advantage most vital to advancing new, disruptive technologies: flexible regulatory frameworks.

The Role of University-Industry R&D Partnerships

The significance of academic R&D initiatives in fostering growth in developing industries is already growing in recognition. Gross domestic expenditure for R&D within higher education in the United States has increased tenfold from 1981–2013, marking a growing interest in investing in R&D projects embarked on by higher education institutions.3 In addition, the United States has also seen a greater proportion of R&D expenditures targeted specifically at experimental R&D initiatives, rather than basic and applied R&D.4 This is significant as the R&D requirements for the AV industry needs both theoretical and practical research approaches, such as experimental methods on real roads and weather conditions, in order to train the programming software of human driving and climate patterns.5 While these trends may point to rising governmental investments in university R&D, federal funding still accounts for roughly just half of the R&D expenditure of the American universities spending the most on R&D. The remainder is supplemented by various sources, like university reserves and coffers, businesses and non-profits, and, most relevant to the case of the development of the AV, industry firms.6

The value of universities for their cities and industry partners has been widely studied, pointing to the significance of knowledge spillovers that may occur through university-industry collaboration.7 Some of these knowledge spillovers may include the clustering of industries, talent, and services as well as the creation of geographically extended collaborative research networks and connections.8 Given the rise of the global knowledge economy, the need for strategic partnerships has increased, transforming the role of the university from a talent hub to an institution that looks further ahead, honing the functional capacities of universities, firms, and economic geographical regions.9 University-firm partnerships are commonplace in many industries, from Microsoft Cisco and the University of Melbourne, to Audi and the University of Munich, SFK Group and Cambridge, IBM and Imperial College London, Nokia and Aalto University, and BP and UC Berkeley.10 The university-industry partnership between the University of Michigan and Chinese investment firm Frontt Capital Management Ltd is one AV-specific example. In order to experiment with self-driving technology in the local transportation conditions of Chinese cities, Frontt Capital is aiming to create an AV testing facility in Shenzhen.11 The firm has invested $27 million in U-M to establish a joint research centre and to help with the construction of a robotics laboratory.12 Frontt Capital will also provide engineering services and consulting fees for U-M researchers to advise Frontt’s AV project.13

Another prominent example of a university-industry AV-related partnership arose following Toyota’s 2015 five-year pledge to spend $50 million on both artificial intelligence and autonomous driving technology, establishing research centres at Stanford University and MIT.14 With an initial focus on intelligent vehicle technology for hybrid AVs, or partially autonomous vehicles, the company has now three established research centers at Stanford University, MIT, and the University of Michigan, with all three functioning under the umbrella of the Toyota Research Institute.15 New projects will be run by Dr. Gill Pratt, founder of the DARPA robotics challenge, which pioneered the initial creation and development of autonomous vehicles.16 The MIT research team is focusing on machine learning or “advanced architectures,” which will let cars perceive, understand, and interpret their surroundings.17 The Stanford research team is interested in semi-autonomous vehicles as well as teaching cars to navigate the outside world alongside human behaviour analysis of both the driver and pedestrians.18 Finally, the University of Michigan will be responsible for fully autonomous vehicles, to which the Toyota Research Institute made a $22 million commitment over 4 years.19 Toyota’s involvement in the research at U-M has been extensive, as it is a founding partner at the school’s Mobility Transformation Center, a public-private research and development initiative for connected and automated vehicles. Toyota’s Collaborative Safety Research Center is also a major sponsor of the University’s Transportation Research Institute, which focuses on advanced safety technologies.20 In fact, the Mobility Transformation Center has a 32-acre mini-city site on campus where researchers test emerging AV technologies.21

Exhibit 1 illustrates the investments, partnerships, and acquisitions that have taken place since Toyota started acting on their interest to develop an AV, corresponding to the various talent and capital flows that were necessary to instigate this process. From blueprint to wheels on pavement, creating an AV is not neatly compacted in nearby geographic locations. It is a complex, interweaving process of flows of talent and capital that partner with, acquire investment in, and collaborate with various industry and educational bodies across the globe. Nevertheless, the very fact that certain transcontinental talent and capital flows have taken place points to certain inherent locational advantages that are present in the AV R&D field.

Toyota’s $50 million investment is not occurring in a vacuum; it reflects a larger, shift in focus towards AVs, and in particular of AV R&D. In 2016, the US government announced plans to invest $4 billion over 10 years into the research of AV technology, along with beginning to work on a national policy in the hopes of accelerating the “development and adoption of safe vehicle automation through real-world pilot projects.”22 The government hopes to work alongside the technology industry and automotive manufacturers in order to test AVs in “designated corridors throughout the country.”23 Explicitly fostering and investing in geographic corridors of innovation, R&D, and industry advancement are of particular interest as the question in focus becomes: where are these corridors being created and why? The answer lies not only in government mandated R&D initiatives, but also in universities, partnerships, and, quite significantly, regulation.

In January 2016, the National Highway Traffic Safety Administration (NHTSA) updated policy to better understand and facilitate the development of autonomous vehicle technology, while proposing industry guidelines to establish “principles of safe operation for fully autonomous vehicles.”24 By 2017, 33 states had introduced legislation regarding AVs. However, that does not necessarily point to regulatory frameworks that allow for the usage or testing of autonomous driving technology, nor does it encompasses foundational bylaws that define, describe, or direct understandings about AVs.25 In various legal shades and across varied conditions, from mandatory technical and institutional conditions for AV use on roads, to limitations to certain pilots, testing projects and studies, autonomous driving functionalities are permitted in the following states: Arkansas, California, Colorado, Florida, Michigan, Nevada, New York, Pennsylvania, North Dakota, Tennessee, Texas, Utah, and Washington D.C.26

Traditional leaders in the automotive industry have optimized their R&D initiatives of AVs accordingly, targeting some of the strongest university talent in some of the most regulatory flexible states.27 Although universities play a vital role in the R&D of innovative and disruptive technologies by attracting investment and partnerships from international firms, the possibility for this global enterprise of flows of capital and talent starts with local regulatory frameworks. Universities must be in compliance with municipal, State, or federal laws in order to carry out the experimental testing and R&D activities inherent to emerging, disruptive technologies.28 Given the uncertainty surrounding automated transportation, firms dealing with AVs emerge as leaders in areas where they are permitted to formally research, test, and develop AV products. In this way, regulation becomes one of many, if not the first, barrier to entry for firms in AV development.

Toyota’s choice to delegate full-AV and semi-AV R&D to the University of Michigan and Stanford University is not accidental. While their respective expertise in the field is unquestioned, the Japanese company is capitalizing on California and Michigan’s permissive regulatory frameworks that allow for vital on-road testing and experimentation with AV technology. They hope that their early adopter status will make them leaders in the new form of economic activity of automated mobility.29 In return, top American universities are acquiring key early-bird experience in AV R&D. In addition, they will continue to attract high profile investment and resources. As academic anchor institutions, universities are influencing the articulation of the physical and talent-oriented space around them, building more facilities, creating more proximate institutes, and further strengthening the talent leadership of certain research institutions and their respective metros in the emerging AV industry.30

Implications of University-Industry AV R&D

The locational advantage of universities in states with flexible regulatory frameworks not only encourages increased investment and cooperation from industry partners, but can also spur more flexible and adaptable academic labour relations. In fact, university involvement in the firm-led development of fully- and semi-AVs is starting to create ripples in the labour force, enabling much labour mobility.31 High skilled academics are being borrowed, traded, and poached by various firms as well as initiating their own partnerships and liaisons between institutions.32 Established companies have been known to acquire smaller, more specialized businesses to rapidly build their research enterprise, as was the case with Toyota, which acquired the entire team of Jaybridge Robotics Inc.33 However, in order to access the talent without “savaging the universities that are doing that research,” Toyota and other companies have embraced a hybrid hiring approach, allowing professors from Stanford, MIT, and U-M to work part time for both the car giant and their respective universities.34

This “Toyota approach” stands in contrast to the tactics taken by Uber. In 2015 the ride-sharing company poached six principal investigators and 34 engineers from Carnegie Mellon’s National Robotics Engineering Center for their own AV R&D project.35 Since then, Carnegie Mellon has attempted to change the structure of their professorship positions, allowing professors to take a leave for business and return, a decision which allows the university to retain a hold on talent.36 Here we see how the transportation revolution is not limited to our roads, but is sending waves through the traditionally inflexible structure of the work force in academia, spurring an opportunity for more flexible and diversely contracted work. Incumbent automotive industries have been forced to adapt following the revolutionary technology of AVs, but universities are also being pushed to adapt. Naturally, as companies target university talent in flexible regulatory frameworks, the evolutions in academic labour relations are suspected to be most concentrated in the same institutions. While we are still in the early stages of AV R&D, as the revolution unfolds with the wide array of industrial, labour, and social ramifications and restructurings, it will be crucially important to track how the structures of our lives evolve in response to this disruptive technology.

University-industry partnerships through the eyes of the labour force are often framed as either ones that opt to partner or sponsor academic talent, or ones that poach individuals from academic institutions. While poaching strategies such as those employed by Uber certainly have their benefits—especially concerning the profit margins of the firm that does the poaching—they also help accelerate innovation, creating access to funding and technology that may otherwise be unavailable. However, partnering with an institution leaves precedent for continued talent growth as innovation, by the very nature of the word, implies continuous development, whereas its counterpart, invention, implies a one-time fix or creation. Even if an individual is temporarily poached by an industry actor for their support in R&D initiatives, no intellectual capital is shared with the students, colleagues, or institution. It simply circulates between the individual and their industry partner. However, in establishing partnered links with institutions—including faculty, staff, and a variety of students and resource management—faculty members not only gain experience in real industry settings, but are also able to remain at the institution and share that knowledge with students and colleagues, perpetuating niche talent. Poaching strategies are made in the mindset of invention: to create, patent, and profit. Those that seek to build on talent and create a precedent for future growth and the development of deeper intellectual faculties, operate on a level of innovation, continuously generating new solutions built on previously laid foundations. Nevertheless, the longer-term benefits of the evolving arena of AV R&D are quite distinctly felt in academic institutions that possess the locational advantage of permissive regulatory frameworks that attract the industry partnership in the first place.

What’s Next?

The AV’s revolutionary potential is not limited to future restructuring of cities and streets; its shockwaves are being felt now, through restructurings in academic labour relations and investment distribution and determination. However, state and industry investments in AV R&D are highly uneven, creating corridors where talent clusters. They will eventually facilitate future clustering of capital from AVs in those same corridors, and concentrate the labour-transformative repercussions of university-industry AV R&D with those same academic institutions targeted with very intentional investment.

The uneven distribution of R&D investment and support is not accidental. It is occurring in areas with open or flexible regulation that allow for early, rigorous testing, piloting, or implementation of disruptive technologies. University talent and knowledge will be hatched first, establishing significant progress and an inevitable head start in a time-sensitive and competitive field. Subsequently, it is then those experienced universities with more established research projects that will catch the attention of investors and firm partners.

When it comes to new and emerging technology, regulation is often the first barrier to establishing success and leadership in the R&D field. As a result, we are seeing states with fast-moving and flexible AV regulatory frameworks, such as California and Michigan, pioneering the development of talent in AV domains, attracting significant international investment and reimagining the role of universities in harnessing economic restructuring. The locations that are capable of performing vital and comprehensive testing, have an established precedent of academic involvement and talent in the domain, and have subsequently garnered significant industry investment and support will emerge as leading clusters of the impending revolution of autonomous mobility.


1 David Ticoll, “Driving Changes: Automated Vehicles in Toronto.” University of Toronto Transportation Research Institute. October 15, 2015.

2 Michele Bertoncello and Dominik Wee, “Ten Ways Autonomous Driving Could Redefine the Automotive World.” McKinsey & Company, 2015.

3 “Research and Development Statistics (RDS),” OECD. Stat, 2016.

4 Ibid.

5 David Ticoll, “Driving Changes: Automated Vehicles in Toronto.” University of Toronto Transportation Research Institute. October 15, 2015.

6 Ronda Britt. “Universities Report Fourth Straight Year of Declining Federal R&D Funding in FY 2015.” National Science Foundation, 2015.

7 Roderick Ponds, Frank van Oort, and Koen Frenken. “Innovation, Spillovers and University–industry Collaboration: An Extended Knowledge Production Function Approach.” Journal of Economic Geography 10 (2010): 231–55. doi:10.1093/jeg/lbp036.

8 Ibid.

9 Ibid.

10 Gail Edmondson, “MAKING INDUSTRY-UNIVERSITY PARTNERSHIPS WORK Lessons from Successful Collaborations.” Science Business Innovation Board, 2012.

11 Nicole Moore Casal, “$27M Investment to Globalize Driverless Vehicle Research.” University of Michigan News, October 15, 2016.

12 Ibid.

13 Ibid.

14 Alex Davies, “Toyota Finally Gets Serious About Self-Driving Cars.” Wired, September 4, 2015.

15 Ibid.

16 Drew Olanoff, “Toyota Pledges $50M To Research AI For Autonomous Vehicles, Hires DARPA’s Dr. Gill Pratt.” TechCrunch, September 4, 2015.

17 Mike Ramsey, “Toyota Teams Up With University of Michigan Researchers on Autonomous Driving.” The Wall Street Journal, April 7, 2016.

18 Alex Davies, “Toyota Finally Gets Serious About Self-Driving Cars.” Wired, September 4, 2015.

19 Laura Lessnau, “Toyota Research Institute Partners with U-M to Accelerate Artificial Intelligence Research.” University of Michigan News, August 10, 2016.

20 Ibid.

21 Ibid.

22 Frederic Lardinois, “US Government Plans To Invest $4B In Autonomous Driving Research Over The Next 10 Years.” TechCrunch, January 14, 2016.

23 Ibid.

24 Amanda Essex and Anne Teigen. “Autonomous Vehicles Legislative Database.” National Conference of State Legislatures (NCLS), June 20, 2017.

25 Ibid.

26 Ibid.

27 Gabriel Weiner and Bryant Walker Smith, Automated Driving: Legislative and Regulatory Action, The Centre for Internet and Society, 2017,

28 Ibid.

29 Ibid.

30 Gail Edmondson, “MAKING INDUSTRY-UNIVERSITY PARTNERSHIPS WORK Lessons from Successful Collaborations.” Science Business Innovation Board, 2012.

31 Roderick Ponds, Frank van Oort, and Koen Frenken. “Innovation, Spillovers and University–industry Collaboration: An Extended Knowledge Production Function Approach.” Journal of Economic Geography 10 (2010): 231–55. doi:10.1093/jeg/lbp036.

32 Mike Ramsey, “Toyota Teams Up With University of Michigan Researchers on Autonomous Driving.” The Wall Street Journal, April 7, 2016.

33 Mike Ramsey, “Toyota Hires Entire Staff of Autonomous-Vehicle Firm.” The Wall Street Journal, March 9, 2016.

34 Mike Ramsey, “Toyota Teams Up With University of Michigan Researchers on Autonomous Driving.” The Wall Street Journal, April 7, 2016.

35 Ibid.

36 Ibid.