Autonomous robot for tunnel mapping

Abstract

We present a novel turnkey system for autonomous tunnel mapping called LiTANK - LiDAR TANK. The system depends entirely on LiDAR sensor readings, without the use of other sensors such as GPS or odometry. Using point cloud registration to perform real-time simultaneous localization and mapping (SLAM) the pose and the surrounding's mapping are estimated. Based on the pose and mapping an autonomous navigation algorithm is applied. Our system was tested inside a hallway with similar characteristics to a tunnel. It navigated autonomously to a selected finishing position with great accuracy and while mapping large unknown area. The final results prove our system concept.