Development of Perilous Environment Estimation System Using a Teleoperated Rescue Robot with On-board LiDAR*

This paper describes a perilous environment estimation system involving the use of a teleoperated rescue robot with on-board LiDAR supported by a human teleoperator to avoid mis-operation and effectively inspect the perilous environment. Teleoperated rescue robots have been deployed recently in actual disaster-stricken areas, and the performance of the rescue robots is also improving. However, some incidents such as the robot being stuck in an obstacle tend to occur. Not only does this prevent the robot from returning, but the stuck robot may also become a new obstacle in the disaster area. In situations involving mis-operations, utilizing failed operational experiences is important when the operator undergoes training related to teleoperation skills, and it is also effective for actual disaster situations. In this paper, we propose a storing method for perilous environments and an estimation method between stored perilous environmental information and current measured information by using a histogram intersection method. In this system, the environmental information is described as a point cloud that is obtained from the LiDAR on-board the robot. Further, using mock-up stairs, the effectiveness of the proposed method in estimating the perilousness of the current environmental information based on the perilous environment as measured from the stored environments is confirmed.