Data-Driven Security Verification for Autonomous Debris Removal in Space

Autonomous Debris Removal allows debris in space to be collected and removed safely from the orbit. However, it relies on the integrity of the sensor data such as LiDAR to estimate the pose of the target (debris). If the sensor data is manipulated, the service vehicle may fail to grasp the target or in worse case damage the vehicle itself or nearby satellites. This paper utilises the multi-modal sensor data to verify the integrity of the sensor data to ensure safe operation of the debris removal. Using the specific domain knowledge for the Autonomous Debris Removal, we propose a computationally efficient algorithm to confirm the point cloud data has not been tampered with. Our experimental results show that significant modification of point cloud can be detected.