Gearing Up For Self-Driving Cars
Fifty years from now, when people relax in automated cars during their commutes and future 16-year-olds laugh at steering wheels the way today’s 16-year-olds laugh at ’90s cell phones, 2016 may be remembered as the year when the development of self-driving cars kicked into high gear. One reason is because of government. Both the Administration and Congress are proactively analyzing the value of self-driving cars and assessing the current state of vehicle regulations. Today, in fact, the U.S. Senate Commerce Committee will be hosting a timely hearing on “The Future of Self-Driving Cars” to obtain more information on development and adoption issues facing self-driving cars.
This is doubly important because Big Auto is starting to really invest in this game. And since companies will almost always be able to innovate faster than regulations change, it is especially important that the government start working now to review regulatory obstacles to self-driving cars. As autonomous car advocates made clear at SXSW this year, a national approach to regulation will ensure quicker adoption and development of the technology, meaning more immediate benefits for safety and for the environment.
Investment Coming from Both Industry and Government
One of the key themes of Project DisCo–our raison d’etre in fact–is highlighting the power of innovation to disrupt industries and promote competition, and what technology did for communications is now being done for the automobile. Exhibit 1: GM’s $1 billion acquisition of Cruise Automation last week. Cruise originally focused on building autopilot features for cars before shifting more ambitiously to developing fully autonomous “kits” which could be retrofitted onto certain cars, specifically the Audi S4 or A4. The acquisition will now make the company’s “full support and resources available” (in the words of GM President Dan Ammann) for developing autonomous software for GM vehicles. And since GM made a record net income of $9.7 billion last year selling a record 9.8 million cars worldwide…well, that’s a whole lotta resources for a whole lotta cars.
For good measure, here’s Exhibit 2: Toyota’s pledge last year to invest $1 billion in advanced technologies like self-driving cars and, last week, its hiring of the entire 16-person staff of Massachusetts-based Jaybridge Robotics.
Two big automakers, two billion-dollar investments, and two very good reasons to believe that self-driving cars will be here sooner than we think. Fortunately, this potential development is being recognized and encouraged here in the U.S. One part of President Obama’s proposed fiscal 2017 budget is the “21st Century Clean Transportation Plan”, an initiative to invest in more resilient, climate-friendly, and innovative transportation systems. This includes a proposal to provide nearly $4 billion in funding over 10 years for developing the autonomous vehicle sector. Similar action is taking place across the Atlantic, where it is believed that this week UK Chancellor George Osborne will propose in the country’s 2016 budget more investments for driverless cars, funds for tests on the country’s roadways, and an overhaul of regulation to get driverless cars on the road by 2020.
Old Regulations Die Hard?
Comparative politics notwithstanding, these funding proposals are a positive indication for government attention to this area (even if Congress doesn’t give President Obama all the money he asks for.) Just as important, these proposals recognize that sometimes the biggest impediments to a government initiative are the very regulations promulgated by government itself.
The automobile industry, well-developed as it is, has to comply with many regulations developed years ago to ensure cars meet safety standards to mitigate the inherent risks of driving. To sell cars in America, companies must certify that their vehicles pass the Federal Motor Vehicle Safety Standards (FMVSS), first introduced in the National Traffic and Motor Vehicle Safety Act of 1966, which mandate the presence of many important things like seatbelts and brakes. Ensuring public safety and keeping highways from looking like scenes from Mad Max: Fury Road are reasonable regulations.
Yet these same regulations could be unduly stifling innovation and thus the mechanisms by which self-driving cars become a (safe) part of reality. To that end, the Department of Transportation’s Volpe National Transportation Systems Center, a sponsor-funded agency focused on research, recently released a preliminary report detailing how standards in the FMVSS could pose barriers to the certification of automated vehicles.
Several standards make reference to a ‘driver’ who has some tangible connection with the vehicle. Of course, that leaves open the consideration of what defines a driver, and to that extent the National Highway Traffic Safety Administration (NHTSA), which provides interpretation of the standards, has shown an openness toward considering new technologies. In a letter to Google, which was concerned about whether its Self-Driving System (SDS) could be interpreted as a driver, NHTSA confirmed that the system could meet the definition, although that does not guarantee that it would then meet all the relevant standards.
However, the report also finds barriers to autonomous vehicle certification in the FMVSS. This includes explicit language such as standard 571.135, section S5.3.1, which states that, “The service brakes shall be activated by means of a foot control.” (emphasis added) One of the caveats in NHTSA’s response to Google was that the SDS would not be able to satisfy this requirement, since in NHTSA’s words: “Given that there will be no foot (or even hand) control to be activated–indeed, given that the SDS will have neither feet nor hands to activate brakes–we understand that Google’s described vehicle design would not comply with S5.3.1 as written.”
This explicit language may unnecessarily limit what self-driving cars can do. An interesting comparison with this “foot control” language is the fact that NHTSA clarified that BMW could deploy a system where the car parks itself while the driver is outside the vehicle by using a special pump to depress the brake so that the car may be shifted out of Park. This is because the relevant standard (571.114, S5.3) does not specify that this action must be done by a foot. Permitting an automated car to use brakes in the normal course of driving, rather than pressing the brake to change transmission, comes down quite simply to whether a regulation specifically mentions feet.
Undertaking studies like these is a good sign that government agencies are aware of their own regulatory obstacles and are making efforts to fully understand the impact of autonomous vehicles. Transportation Secretary Anthony Foxx announced earlier this year that NHTSA will be working with industry and state stakeholders to test self-driving cars and help develop a model policy that could be applied across states. On April 8th, the organization will hold a public hearing in Washington, DC to gather input for developing guidelines. And one of the reasons for the Senate Commerce Committee’s hearing today is to solicit views on “the appropriate role of government in promoting innovation including removing unnecessary hurdles…”
Hopefully, this hearing and the Administration’s actions are an indication that the federal government can work as a team to promote the adoption of self-driving cars. They will need to, since the speed of innovation time and time again outpaces the regulatory process. And if GM’s bid is any indication, the race to introduce these cars to the market is only going to intensify. Indeed, GM’s president says so himself: “This [acquisition] will significantly accelerate the timeline for bringing autonomous vehicles to market.”
By taking early action, recognizing the potential of the technology, and mobilizing shareholders, governments may just be able to keep pace with the speed of innovation.
Jordan Harriman is a Policy Fellow at CCIA.