Today, regulators are shining the spotlight on autonomous vehicles and the robotaxi. In California, for example, legislation has been introduced in the state Senate that would allow city and county governments to impose stricter limits and regulations on robotaxi services than those handed down by state agencies.
Senate Bill 915 comes after a string of high-profile autonomous vehicle accidents involving Cruise driverless vehicles in San Francisco last year, including a collision with a fire truck and an AV rolling over a pedestrian who’d been struck by another car. California’s Department of Motor Vehicles shut down Cruise’s operations statewide in October, and the General Motors subsidiary now says it is refocusing on “trust, accountability and transparency.”
But autonomous vehicles are still out there. Waymo, Alphabet’s AV subsidiary, is in four cities, including San Francisco. It’s giving a road show “tour” across Los Angeles. In Phoenix, it has partnered with Uber to give riders a driverless option.
There is hope that robotaxis and other driverless vehicles may one day improve road safety and mobility writ large. But in the short term, how should city and state agencies — and companies like Cruise and Waymo — deal with the risks they pose today and earn that public trust? How can the sometimes-conflicting interests of all stakeholders be considered and balanced?
It’s a complicated problem. Robotaxis are a tangle of technologies and services. A robotaxi marries an autonomous vehicle with mobility on demand, and a service may or may not employ electric vehicles. This has created ambiguity as to who has authority to oversee or regulate what. In California, for example, the DMV issues permits to AVs and regulates vehicle safety, the Public Utilities Commission regulates commercial passenger services and is focused on passenger safety, and the Department of Transportation handles road infrastructure and highways.
SB 915 aims to empower local governments to ensure public safety and accountability, but matters are further complicated by the state’s patchwork regulatory regime and geographical reality: Roads can cross multiple city boundaries, creating complex legal frameworks and issues with determining liability when things go wrong.
Along with state agencies and local leaders, add to the decision-making mix the AV companies and angry local residents, and that’s a lot of stakeholders who don’t necessarily agree on the benefits or impacts of robotaxis on society, the economy, the environment or road safety. Even more than e-scooters and ride-sharing before them, robotaxis are a novel mode of transportation; there’s little experience to draw upon when figuring out what the rules and regulations should be.
There is, however, a process that has been designed to deal with such messy situations. It’s called “decision-making under deep uncertainty,” or DMDU. Other industries, such as water and energy, have already successfully used DMDU methods, but it’s still fairly new to the transportation sector.
DMDU focuses on actions that can be agreed upon, even if multiple stakeholders agree on little else. For instance, instead of speculating about what does or doesn’t make robotaxis unsafe, DMDU could help regulators and AV companies figure out which warning signs to monitor. In the case of robotaxis, that might be crashes per vehicle miles of travel compared with other mobility services operating in the same area, such as Uber or Lyft. Such an indicator can evolve as more data comes in and the technology becomes better understood. But stakeholders aren’t paralyzed in the meantime; they can use the warning signs to adapt before something goes terribly wrong.
When data is not available, agencies can use DMDU-inspired methods to measure proxies for risk, such as how rapidly the technology is adopted, whether an agency has the ability to manage the risk, or the size of the potentially affected population. In the case of robotaxis, such a list might also include the number of vehicles deployed, number of trips completed, number of users, and traffic density in the area of operation.
Even more new transportation technologies are on the way. Sidewalk delivery bots may already be bringing food to your neighbor’s door. And some suggest that electric air taxis might be operational in Los Angeles just in time for the 2028 Olympics.
Cities and regulatory agencies can often feel that they’re left as bystanders when new technology is rolled out, yet they are responsible for the safety of whatever private entities develop and deploy. A DMDU-based risk management framework could help them come together to better manage the transitions to new forms of transportation.
Hye Min Park is an assistant policy researcher at the RAND Corp. and a Ph.D. student at the Pardee RAND Graduate School. Her research focuses on energy policy and decision-making under deep uncertainty (DMDU). Fabian E. Villalobos is an engineer at RAND and a professor of policy analysis at the Pardee RAND Graduate School. His research focuses on the intersection of technology, economics and geopolitics. You can follow him on Twitter and LinkedIn.
Governing’s opinion columns reflect the views of their authors and not necessarily those of Governing’s editors or management.
Related Articles
