Reinforcement Learning For Testing Interdependent Requirements In Autonomous Vehicles: An Empirical Study

Abstract

arXiv:2502.15792v2 Announce Type: replace-cross Abstract: Autonomous vehicles (AVs) make driving decisions without humans, making dependability assurance critical. Scenario-based testing is widely used to evaluate AVs under diverse conditions, with reinforcement learning (RL) generating test scenarios that identify violations of functional and safety requirements. Many requirements are interdependent and involve trade-offs, making it unclear whether single-objective RL (SORL), which combines objectives into a single reward, can reliably reveal violations or whether multi-objective RL (MORL), which explicitly considers multiple objectives, is necessary. We present an empirical evaluation comparing SORL and MORL for generating critical scenarios that simultaneously test interdependent requirements using an end-to-end AV controller and high-fidelity simulator. Results suggest that MORL and SORL differ mainly in how violations occur, while showing comparable effectiveness in many cases. M

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keywu2026reinforcement

Related papers

• Reinforcement Learning For Online Testing Of Autonomous Driving Systems: A Replication And Extension Study (2024)4.52
• Multi-agent Uncertainty-aware Pessimistic Model-based Reinforcement Learning For Connected Autonomous Vehicles (2025)0.00
• Rigorous Agent Evaluation: An Adversarial Approach To Uncover Catastrophic Failures (2018)0.00
• Assured Learning-enabled Autonomy: A Metacognitive Reinforcement Learning Framework (2021)0.00
• On Assessing The Safety Of Reinforcement Learning Algorithms Using Formal Methods (2021)0.00
• Test-driven Reinforcement Learning In Continuous Control (2025)0.00
• Continuous Coordination As A Realistic Scenario For Lifelong Learning (2021)0.00
• On Generalization Across Environments In Multi-objective Reinforcement Learning (2025)0.00