Scholarlink

An Automated Driving System (ADS) may encounter infinite scenarios in the real world, making it impossible to test every scenario in simulation. To address this, we previously introduced the Simulation-Based Safety Testing Scenario Selection (SSTSS) process, which prioritizes and selects scenarios for simulation-based safety testing of ADS. In this paper, we select the top-prioritized scenario identified by the SSTSS process to evaluate the safety behavior of UT-ADS, an ADS developed at the University of Tartu. We simulate a “Follow Lead Vehicle” scenario to evaluate if UT-ADS maintains a safe following distance as per the Responsibility-Sensitive Safety (RSS) model. Our preliminary simulation results show that UT-ADS maintains safe following distances at lower speeds but violates the RSS minimum safe distances at higher speeds. Our simulation results could provide useful feedback for UT-ADS developers.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

From Scenario Selection to Simulation: Safety Testing of an Automated Driving System

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Fauzia Khan,
  • Ali Ihsan Gullu,
  • Hina Anwar,
  • Dietmar Pfahl

摘要

An Automated Driving System (ADS) may encounter infinite scenarios in the real world, making it impossible to test every scenario in simulation. To address this, we previously introduced the Simulation-Based Safety Testing Scenario Selection (SSTSS) process, which prioritizes and selects scenarios for simulation-based safety testing of ADS. In this paper, we select the top-prioritized scenario identified by the SSTSS process to evaluate the safety behavior of UT-ADS, an ADS developed at the University of Tartu. We simulate a “Follow Lead Vehicle” scenario to evaluate if UT-ADS maintains a safe following distance as per the Responsibility-Sensitive Safety (RSS) model. Our preliminary simulation results show that UT-ADS maintains safe following distances at lower speeds but violates the RSS minimum safe distances at higher speeds. Our simulation results could provide useful feedback for UT-ADS developers.