Bian, Jiang (Central South University), Zhou, Rui (central south university), Huang, Helai (Central South University)

Accelerating Autonomous Vehicle Safety Testing: A Framework Balancing Scenario Risk and Test Coverage

Scheduled for presentation during the Regular Session "S12b-Safety and Risk Assessment for Autonomous Driving Systems" (WE-EA-T12), Wednesday, November 19, 2025, 13:50−14:10, Broadbeach 3

2025 IEEE 28th International Conference on Intelligent Transportation Systems (ITSC), November 18-21, 2025, Gold Coast, Australia

This information is tentative and subject to change. Compiled on October 19, 2025

Abstract

Scenario-based testing remains a primary approach for validating autonomous vehicle safety, yet the exhaustive evaluation of all generated scenarios is highly resource-demanding. To improve efficiency, we introduce a scenario selection framework that jointly considers scenario risk and coverage. By analyzing previously tested cases within the operational design domain (ODD), the framework infers the likely distribution of untested high-risk scenarios from observed crash patterns. This insight informs the selection of scenarios for subsequent simulation rounds. Experimental results using Baidu Apollo show that the selected scenarios were both representative and effective: 82.2% resulted in unavoidable collisions, while the remainder displayed elevated risk levels as measured by surrogate safety indicators such as post-encroachment time (PET) and generalized time to collision (GTTC).