Today at the North American International Auto Show in Detroit, Transportation Secretary Anthony Foxx announced new initiatives on automated vehicles, commonly referred to as driverless cars, self-driving cars, or autonomous vehicles. Foxx, joined by representatives from Google, Tesla, the Big 3 automakers, and Tier 1 supplier Delphi, discussed the Obama administration’s plan to pump $4 billion over 10 years to “accelerate” the development and deployment of automated vehicles. How they exactly plan to do so—and to ensure these requested funds won’t be squandered on pet projects or on initiatives that delay development and deployment—remains to be seen.
- Within six months, NHTSA will propose guidance to industry on establishing principles of safe operation for fully autonomous vehicles (vehicles at Level 4 on the scale established in NHTSA’s 2013 preliminary policy statement).
- Within six months, NHTSA, after working with the states, will propose model policy for the states to adopt aimed at avoiding state-to-state conflicts and preserving NHTSA’s position as the one-stop national auto safety regulator.
- NHTSA will work with automakers and suppliers to promote regulatory flexibility by making use of its interpretive and exemption powers.
Of course, DOT/NHTSA couldn’t let this go without throwing in something completely unrelated to automated vehicle technology—connected vehicle technology promoted most heavily by interest groups like the Intelligent Transportation Society of America and Japanese automakers—by noting that it will continue “to incentivize the development and adoption of technologies using vehicle-to-vehicle and vehicle-to-infrastructure communications, so that Americans enjoy the full benefits of connected-vehicle safety technology.” For more on why NHTSA’s approach to connected vehicle technology is anti-technology, anti-safety, and anti-taxpayer, see here.
For the last few years, we at the Competitive Enterprise Institute (CEI) have advocated a number of policy reforms to help usher in wider testing and ultimate deployment of automated vehicles. In April 2014, we released the first free-market policy whitepaper on the technology, “Self-Driving Regulation: Pro-Market Policies Key to Automated Vehicle Innovation.”
In the paper, we urged policymakers to adopt an optimistic but hands-off stance, arguing that automated vehicles can be best promoted not by heavy-handed government technology mandates or bloated federal research programs, but by allowing the private sector to continue to lead on development. This was meant to serve as a stark contrast with the approach taken by states such as California, which moved too quickly to enact legislation and are now attempting to implement regulation in a way that won’t cause developers to flee their states. Unfortunately, this is proving to be extremely difficult and it is becoming clear that it may not be possible under the current statutory frameworks.
Although CEI’s 2014 paper expressed general skepticism of passing legislation and promulgating regulations on automated vehicles at this early stage, we did provide specific recommendations in the following five policy areas:
- Legality. States should enact policies that explicitly recognize the legality of automated vehicles, but should do so without imposing technology-specific mandates. The American Legislative Exchange Council adopted a model resolution that lays out these principles in January 2014.
- Safety. Policymakers, particularly members of Congress and officials at the National Highway Traffic Safety Administration, must examine current regulations and determine which may present conflicts with automated vehicle technology. For instance, Federal Motor Vehicle Safety Standard 108 defines the “vehicular hazard warning signal operating unit” (the switch for emergency flashers) as “driver controlled.” University of South Carolina law professor Bryant Walker Smith has noted that this would possibly preclude automated engagement (pp. 459-60). But even if there are few direct conflicts, a number NHTSA’s vehicle safety standards when applied to automated vehicles will produce far fewer benefits once the human driver is taken out of the equation. As a result, safety standards that passed benefit-cost analysis when a human was assumed to be driving a vehicle are likely to fail it when computers take over many and perhaps all driving tasks previously handled by humans. When such rules as applied to automated vehicles fail benefit-cost analyses, policymakers should be prepared to discard them or at the very least exempt automated vehicles.
- Infrastructure. Policymakers should reject mandated vehicle-to-vehicle (V2V) and/or vehicle-to-infrastructure (V2I) communications technologies as necessary to highway vehicle automation. As CEI noted in more detail in an October 2014 proceeding before NHTSA on V2V, a number of technical challenges make selecting an appropriate V2V/V2I standard that could actually promote safety and mobility exceedingly difficult, let alone facilitating automated vehicles. Further, the massive infrastructure investments necessitated by a large scale V2I deployment strategy are not possible in a constrained funding environment. Filling potholes is proving hard enough; deploying roadside V2I units every 300 meters of urban expressway at more than $50,000 a pop is a nonstarter.
- Products liability/insurance. Despite some pundits’ claims to the contrary, there is no indication that products liability and insurance present insurmountable hurdles. Indeed, insurers have been active and early participants in the automated vehicle research community. To be sure, automated vehicles may change the automotive risk landscape by raising manufacturers’ duty of care due to the heightened risk foreseeability in a data-rich environment and shifting insurance product coverage away from human “operators” and toward manufacturers, but why would we expect automotive products liability and insurance to remain the same when humans are no longer directing vehicles?
- Transportation services. Think Uber, but with automated vehicles. Robotaxis are often touted in the press as the most transformative application of automated vehicles. Indeed, this is an area of much promise that could greatly reduce the cost of not owning an automobile, particularly for those in dense urban cores. As states and municipalities around the country have been busy carving out exemptions for Uber and its extremely similar competitors, few are deciding to take a sledgehammer to the longstanding anticompetitive transportation service regulations. This presents a problem, as many of the “transportation network company” (a term of art coined by California regulators in 2013 that is in fashion among politicians to unartfully describe for-profit e-hailing platforms) carve-outs assume a human driver is in control of the vehicle—indeed, most of the fighting between right-wing (and Uber et al.) and left-wing (and unions) over e-hailing platforms centers on compensation, background checks, and other driver-exclusive topics. As I warned in July 2014, failing to dismantle the existing anticompetitive transportation service regulations could well result in e-hailing platforms taking the place of the taxi cartels in resisting change in a future where human drivers become obsolete.
Today’s announcement is a welcome sign from the Obama administration that they take automated vehicle technology seriously. But supporters of a driverless future must remain vigilant. As we’ve said before, the greatest threat to automated vehicles comes from well-intentioned politicians and bureaucrats imposing their worldviews on a nascent and evolving technology. If automated vehicles are indeed demonstrated to be safer than traditional manually driven ones, political missteps that delay the consumer release or raise the price of automated vehicles will translate to increased property damage, injuries, and fatalities.