Autonomous vehicles (AVs) have already driven millions of miles on public roads, but even the simplest maneuvers such as a lane change or vehicle overtake have not been certified for safety. Current methodologies for testing and verification of Advanced Driver Assistance Systems such as Adaptive Cruise Control cannot be directly applied to determine AV safety as the AV actively makes decisions using its perception, planning and control systems for both longitudinal and lateral motion. These systems increasingly use machine learning for which it is fundamentally hard to derive safety guarantees across a range of driving scenarios and environmental conditions. New approaches are needed to bound and minimize the risk of AVs to assure the public, determine liability and insurance pricing and ensure the long-term growth of the domain.
This talk will describe the investigations in the DoT Mobility21 National University Transportation center at the University of Pennsylvania, with a focus on 3 questions:
- What type of evidence should we require before giving a driver’s license to an autonomous vehicle?
- How do we build a search engine to efficiently capture unsafe driving instances through AV safety benchmarks before the vehicle hits the road?
- How can automotive OEMs reason about Robot Safety Laws across all driving scenarios?
This talk will show lots of AV crashes and explore testing and verification in building safe autonomous vehicles.