Towards LIDAR-RADAR based terrain mapping

Abstract

This paper addresses the problem of perception for autonomous vehicle navigation in real environments. Integrity safe navigation of autonomous vehicles in unknown environments poses a traversability problem. We are interested in the integrity-safe navigation in unknown environments. Safe navigation is a task that depends on the knowledge of the surrounding environment and the vehicle dynamics. Classical navigation approach focus on obstacle avoidance often based on occupancy and elevation maps. We propose to combine an optical sensor and an electromagnetic sensor to build a richer map of the environment which will be used for traversability analysis and path planning. The proposed lidar-radar map encodes the geometry of the environment such that traversability analysis and trajectory planning guarantee the robot's integrity in a stability sense. A comparative analysis of two mapping algorithms using lidar, radar, IMU and GPS sensors shows the advantages of such bimodal perception system. Results have been validated experimentally.