Autonomous vehicles (AV) provide the opportunity to correct government transportation failures is how the Brookings Institution’s Clifford Winston characterized the potential opportunity provided by autonomous vehicles. Winston spoke to the possible economic impact of autonomous vehicles in an online media briefing (YouTube video) that also included speakers from Princeton and the Reason Foundation who touched on the technology and the role of public policy and regulation. A lively question and answer period followed the briefing.
Vehicle Automation Means Many Things #
Princeton’s Alain Kornhauser kicked off the event with a simplified definition of NHTSA’s six levels (PDF) of automation into three categories
- Safe Driving (NHSTA Level 0 to 2) – Advanced Driver Safety Systems, such as anti-lock brakes, override the driver to avoid a crash.
- Advanced Driver Assistance Systems (NHSTA Level 3) – ADAS are systems, like intelligent cruise control and automatic lane keeping, that allow the driver to shift roles to that of a monitor. With that said, the human behind the wheel is still responsible and must be ready to take control as the driver at any time. Kornhauser characterizes these features as “comfort and convenience” for the driver.
- Driverless (NHSTA Levels 4 & 5) – Both Level 4 and Level 5 mean traditional human controls (e.g., steering wheel, pedals, etc.) are no longer necessary. The distinction is that Level 4 has an operational design domain (i.e., limits such as location, weather conditions, etc.) where it can operate, while Level 5 amounts to full automation and could drive wherever a human could drive (e.g., off-road, etc.).
Kornhauser, who has been involved with transportation automation for more than 50 years, says that full automation implied in the NHSTA Level 5 category is very difficult and not on the horizon.
He also believes that Level 4 will not be good enough for ride-hailing services, as the geographic reach of these services is too large. He thinks a fleet model, like what he is advocating with Trenton Moves, is the way that Level 4 will succeed. In the fleet model, the focus is on providing cost-effective, equitable, and sustainable mobility to an underserved geographic area.
AVs: One of the Most Important Transportation Innovations #
Winston suggests that autonomous vehicles will provide significant economic and quality-of-life improvements. Co-author of the book, Autonomous Vehicles: The Road to Economic Growth?, his calculations indicate as much as a 1% increase in GDP when they get to 50% market share.
He listed more than ten areas where autonomous vehicles have the potential to positively impact society including:
- Reduction in emissions, particularly when combined with electrification
- Economic efficiencies including lower-cost freight, fewer subsidies for public transportation, and reduction in hospitalizations for people in vehicle crashes
- Improved mobility for those without good mobility options, and faster travel times (he indicates one-third of delays are from incidents – autonomous vehicles do not rubberneck).
With a shared autonomous vehicle model, there is also an opportunity to improve land use by reducing parking requirements. He contends that road congestion pricing will be a necessary tool to minimize sprawl that AVs could potentially spark. At the same time, AVs have the potential for greater throughput on existing roads than human-driven cars.
It is disappointing to him that economists and others do not have more positive and constructive thinking regarding how to integrate AVs into society. He reminds everyone to be patient as it will be a long time before autonomous vehicles become mainstream.
He points out that the technology may be secondary to policy in terms of getting AVs on the road. Winston states, “The obstacle to AV’s success in the U.S. is government policy.”
The Policy Challenge #
The Reason Foundation’s Marc Scribner gave a high-level, but detailed overview of the complex policy web that stretches across levels of government and their associated agencies. The federal government has primary authority over vehicle safety and performance. At the same time, regulation of operations and infrastructure is within the respective domains of state and local governments.
Underpinning federal regulations is the 1966 National Traffic and Motor Vehicle Safety Act. This Act relies on the self-certification of vehicles by the manufacturers. This also has led to 73 mandated and 255 voluntary consensus standards with a median age of these regulations dating to 1980 according to Scribner.
Further, the average NHSTA rulemaking to change a rule is approximately 8 years, so it will take a while for regulations to catch up to technology and/or business model changes, even if the rule is for a safety improvement and not just convenience and comfort. As an example, it took 9-years for NHSTA to change regulations to permit headlights that turn with a vehicle.
Scribner cites the March 30th, 2022 NHSTA ruling as clearing the way for self-certification of occupant-less cargo vehicles. This ruling will be helpful for robotic delivery vehicles, such as those from Nuro, but Scribner suggests there is still much work to be done in the regulatory and policy domain before autonomous vehicles are ubiquitous.
From a state perspective, forty-two states have enacted autonomous vehicle bills. These focus on operational topics, such as vehicle code updates, ride-hailing integration, first responder integration, and policies to support hub-to-hub transfers of autonomous trucking. Congress is the wildcard, as to whether there will be stand-alone, AV national legislation or whether legislators will attach specific AV rules to must-pass legislation.
Meanwhile, back in the Year 2040
The discussion and the following question and answer session hearken back to the late 1990s and the early days of the Internet. Sure, the basic technology to bring the Internet to the home worked, but it took time for all the elements to come together to enable profitable, web-based use- cases. It feels like the autonomous vehicle industry is following a similar path.
Leave a Reply