Active Terrain Relative Navigation with Surface Mapping Sensors

This presentation details the development of an Active Terrain Relative Navigation (ATRN) system as part of JSC's Center Innovation Fund program. This research focuses on creating rotationally and scale invariant 3D feature identification algorithms for precise planetary landing and surface navigation applications in any lighting condition. The ATRN system demonstrates impressive accuracy with 3D keypoint localization errors of less than 10 centimeters and pose estimation errors less than of 10 centimeters and 0.05° (well below the 1m/1° goals). The technology includes a modular and computationally efficient pipeline that determines absolute vehicle position/attitude using reference maps, calculates relative position between successive scans, and constructs high-resolution terrain maps of observed areas. A standalone C++ library is under development (expected October 2025), with plans for performance evaluation using digital elevation maps (DEMs) obtained during helicopter flight tests from the Goddard Hazard Detection LiDAR at KSC's hazard field. The technology shows promising applications for planetary precision landing and ground surface navigation missions.