Host: The Japan Society of Mechanical Engineers
Name : [in Japanese]
Date : June 28, 2023 - July 01, 2023
In a previous report, we developed a road obstacle detection technique for crawler-type robots using 2D-LiDAR, but experimental results showed that the technique could not detect road obstacles in front of the robot. Based on this problem, we proposed a detection technique using 3D-LiDAR and demonstrated its effectiveness through experiments. Specifically, this technique acquires point cloud data of a vibration-resistant road surface by vibration compensation with plane detection, converts it into a height map that maintains the maximum height of the point cloud in a cell area, and determines obstacles based on the height difference in the height map. Experiments using the proposed method showed that the robot was able to stop in all scenes, confirming that the proposed method improves the performance of obstacle detection.
