Minimizing Safety Interference for Safe and Comfortable Automated Driving with Distributional Reinforcement Learning

Despite recent advances in reinforcement learning (RL), its application in safety critical domains like autonomous vehicles is still challenging. Although penalizing RL agents for risky situations can help to learn safe policies, it may also lead to highly conservative behavior. In this paper, we propose a distributional RL framework in order to learn adaptive policies which allow to tune their level of conservativity at run-time based on the desired comfort and utility. Using a proactive safety verification approach, the proposed framework can guarantee that actions generated from RL are failsafe according to the worst-case assumptions. Concurrently, the policy is encouraged to minimize safety interference and generate more comfortable behavior. We trained and evaluated the proposed approach and baseline policies using a high level simulator with a variety of randomized scenarios including several corner cases which rarely happen in reality but are very crucial. In light of our experiments, the behavior of policies learned using distributional RL is adaptive at run-time and robust to the environment uncertainty. Quantitatively, the learned distributional RL agent reduces the average driving time more than 50% compared to the normal DQN policy. It also requires 83% less safety interference compared to the rule-based policy while only slightly increasing the average driving time. We also study sensitivity of the learned policy in environments with higher perception noise and show that our algorithm learns policies that can still drive reliable when the perception noise is two times higher than in the training configuration in automated merging and crossing at occluded intersections.