TechCrunch Mobility: How do we know when an AV is safe enough?

Welcome back to TechCrunch Mobility, your hub for the future of transportation and now, more than ever, the role AI is playing in it. To get this in your inbox, sign up here for free — just click TechCrunch Mobility! One of the stickier problems for autonomous vehicle developers, regulators, and the public is determining when self-driving technology is safe enough for the road. So far, that question has largely fallen to industry to answer — companies that are applying their autonomous vehicle technology to self-driving trucks, robotaxis, and maybe someday personally owned passenger vehicles. Today, there is no standardized test for judging how well an automated driving system performs in the real world.

We require people to pass a driver’s test before getting a license, but there isn’t an equivalent test or system for autonomous vehicle technology. So how do we know whether an AV is a safe driver? The National Highway Traffic Safety Administration says it’s working on an answer. Federal motor vehicle safety standards (FMVSS) already govern the vehicle automakers sell today. But those FMVSS rules weren’t designed for vehicles like the Tesla Cybercab and Zoox robotaxi, which lack traditional, and federally required, equipment like pedals, steering wheels, and design features like side mirrors and rearview mirrors. The NHTSA is working on a federal automated vehicle framework