A Local Planner Based on Environmental Geometric Features for Rescue Robots

Abstract

For robots that perform autonomous exploration work in unknown environments, local planning is a key technol-ogy to determine whether the robot can work safely. In this paper, we propose a local planner based on the geometric features of the 3D point cloud for rescue robots, where the raw 3D point cloud are divided into Passable Areas (PA), Surmountable Obstacles (SO) and Insurmountable Obstacles (IO). The robot will obtain the orientation information of SO in real time. In the process of autonomous exploration, in order to ensure the safety of the robot, the robot should operate in the passable areas as much as possible. When there is no way, the robot can cross SO with lower risk according to its own obstacle crossing ability. This local planner newly defines the obstacles that can be crossed, so the robot has more flexible choices in the exploration process. The experimental results show that the safety can be improved for rescue robots during autonomous exploration.