Autonomous driving improves safety and efficiency at busy intersections
A Risk-Sensitive and Uncertainty-Aware Decision-Making and Control Framework for Safe and Robust Autonomous Driving
Robotics
Summary
Autonomous cars need to make safe decisions even in tricky traffic spots like busy intersections without traffic lights. The authors created a system that helps self-driving cars better understand risks and how sure they are about decisions. This system adjusts safety rules depending on uncertainty and fixes model errors to keep driving smooth and safe. Tests in simulated intersections showed the system keeps cars safer and more efficient than some earlier methods.
What this means in practice
- •For autonomous vehicle engineers: Enhance decision algorithms in self-driving cars for safer and more efficient navigation at unsignalized urban intersections.
- •For robotics control teams: Implement uncertainty-aware safety mechanisms in robot controllers to balance risk and performance in dynamic environments.
Authors
Zhuoren Li, Ran Yu, Weiqi Zhang, Ming Liu, Lu Xiong, Chen Sun, Bo Leng
Abstract
Reinforcement learning (RL) has demonstrated considerable potential for autonomous driving decision-making. However, its deployment in urban autonomous driving, particularly at highly interactive unsignalized intersections, remains challenging, as learned policies may struggle to maintain both safety and robust decision-making in complex traffic situations. Conventional safety-filtering approaches typically employ fixed conservative constraints, which may improve safety at the cost of excessive intervention and degraded traffic efficiency. To address these limitations, we propose a Risk-sensitive and Uncertainty-aware Decision-making and Control (RUDC) framework for safe and robust autonomous driving. RUDC couples risk-sensitive distributional RL with ensemble-based policy uncertainty quantification, jointly accounting for tail risks in return distributions and uncertainty in learned policies. An uncertainty-aware high-order control barrier function (HOCBF)-based safety correction mechanism adaptively adjusts constraint strictness according to policy uncertainty, while a learnable residual predictor compensates for CBF model mismatches and discretization errors. Extensive simulations at unsignalized intersections demonstrate that RUDC achieves a favorable balance among safety, efficiency, and robustness, outperforming representative safe RL baselines under both nominal and challenging OOD and long-tail scenarios while satisfying real-time requirements.