Worst-Case Hidden-Vehicle Trajectory Search in Spatiotemporal Occlusion Regions
Finding the worst-case hidden drivers that self-driving cars can't see
Autonomous vehicles often lose sight of other cars behind buildings or parked vehicles, forcing them to guess what might be hiding there. Researchers developed a method that systematically searches for the most dangerous hidden-vehicle scenarios—ones that could cause collisions even if the self-driving car responds optimally. Testing on real driving data, the method found six avoidable crashes that simpler prediction methods would have missed.
Self-driving cars currently rely on guessing what hidden vehicles might do, which leaves dangerous edge cases undetected. This method identifies genuine worst-case threats before they happen, allowing engineers to either redesign car behavior or flag scenarios where human takeover is needed. For safety-critical deployment, catching these six types of failure modes before a car hits the road could prevent real collisions.