Interpretable Safety Validation for Autonomous Vehicles
This work addresses the safety validation challenge for autonomous vehicles by providing interpretable failure cases, which is incremental as it builds on existing testing methods to improve clarity and relevance.
The paper tackles the problem of validating autonomous vehicle safety in simulation by developing a method to find interpretable failures described by human-understandable signal temporal logic expressions, which are optimized for high likelihood. The approach demonstrated in unprotected left turn and pedestrian crosswalk scenarios finds more failures with higher likelihood than a baseline importance sampling method while maintaining interpretability.
An open problem for autonomous driving is how to validate the safety of an autonomous vehicle in simulation. Automated testing procedures can find failures of an autonomous system but these failures may be difficult to interpret due to their high dimensionality and may be so unlikely as to not be important. This work describes an approach for finding interpretable failures of an autonomous system. The failures are described by signal temporal logic expressions that can be understood by a human, and are optimized to produce failures that have high likelihood. Our methodology is demonstrated for the safety validation of an autonomous vehicle in the context of an unprotected left turn and a crosswalk with a pedestrian. Compared to a baseline importance sampling approach, our methodology finds more failures with higher likelihood while retaining interpretability.