LIDAR-based Autonomous Pose Determination For a Large Space Debris

Recent NASA simulations regarding the debris population in low Earth orbit have demonstrated the need for the active removal of at least five large objects per year to prevent the triggering of the "Kessler syndrome". However, active debris removal missions pose many significant technological challenges, starting from the rendezvous with an uncooperative target, which is not arranged to be approached by a removal system. In this respect, a major task is to develop reliable and robust techniques for the autonomous determination of the target pose. The aim of this paper is to investigate the performance of a LIDAR-based system for pose determination of a known large debris. LIDAR measurements consist of a 3D-point cloud, so the attention is focused on 3D techniques for pose acquisition and tracking. For pose acquisition, an on-line 3D Template Matching technique is introduced specifically thought for on-board autonomous operations. For pose tracking, different variants of the Iterative Closest Point algorithm are investigated. Specifically, two approaches, namely nearest neighbor and normal shooting, are compared to identify the best trade-off between accuracy and computational resources. To this end, a simulator is developed in which realistic debris geometry and motion is implemented, as well as a safe relative trajectory around the debris, and LIDAR operation. Results demonstrate the effectiveness of the proposed techniques for pose acquisition and tracking.