Sensor Limitation Adaptive Any-Time Planning Based on Trajectory Sampling in Safety Verified Region
Eijiro Takeuchi, Simon Thompson, Shinpei Kato · 2024
This paper introduces a sensor-based trajectory planning method for autonomous vehicles, specifically designed to adapt to sensor limitations such as sensing range, resolution, and occlusions. Autonomous vehicles rely on sensors to detect obstacles and avoid collisions. While the typical planning methods use detected obstacle information to find a safe path to the destination, sensors have limitations in detecting obstacles, and these sensing limitations directly affect the safety of planning results. In contrast, the proposed planner in this paper utilizes safety-verified regions to determine the path. In this method, the planner uses sensors to verify safe regions to traverse, ensuring that sensing limitations do not lead to risky situations. This paper proposes a sampling-based trajectory planning method in safety verified space and illustrates how the method can generate trajectories that adapt to sensing limitations, including sensing range and occlusion. To demonstrate the benefit of this approach, the proposed method is compared with current autonomous driving system. The experimental results are presented for navigation scenarios including, limited sensing range, resolution, obstacle avoidance with occlusion, and moving obstacles with occlusion.