Matthew Wansley (Benjamin N. Cardozo School of Law) has posted Backroom Drivers (Harvard Journal of Law & Technology, forthcoming) on SSRN. Here is the abstract:
We increasingly share the roads with vehicles that have no driver behind the wheel. But some of these vehicles still have a human in the loop. Today’s automated driving systems (ADSs) rely on help from remote facilitators who can affect how a vehicle moves without touching its controls. Remote facilitators typically assist the ADS by providing information and advice. Some can brake, accelerate, and steer in real time. Either way, their role is safety-critical. Remote facilitators can help ADSs navigate edge cases like construction, police activity, and extreme weather. But they also introduce new risks. Latency or dropouts can disrupt the communication link. A surge in requests for help can overwhelm the system. And the human in the loop can misapprehend the vehicle’s situation or become fatigued. In this Essay, I consider how we should regulate remote facilitation. I argue that regulators should aim to surface unreasonable risks proactively while giving companies flexibility in how they manage those risks. I propose adversarial audits, emergency simulations, and emergency geofencing.
Recommended!
To receive new posts from Legal Theory Blog by email, get a free subscription to Legal Theory Stack.
Lawrence Solum