The first collection of multi-modal hardware-in-the-loop data is underway at the Testbed for Rendezvous and Optical Navigation (TRON) by a joint AcademiaIndustry team within the Center for AEroSpace Autonomy Research (CAESAR) composed of Pol Francesch Huc, Emily Bates, Tae Ha "Jeff" Park from Stanford Space Rendezvous Laboratory (SLAB) and Kuldeep Barad from Redwire Space Europe.
After a smooth H/W + S/W integration phase, the team conducted a comprehensive workspace calibration for precise data annotation of stereo imagery. The team is now generating a new machine learning dataset for diverse Rendezvous and Proximity Operations (RPO) trajectories, powered by SLAB's relative astrodynamics simulator.
The goal is to understand the challenges of working with real multimodal sensors for space applications and develop practical on-board solutions for characterization of unknown spacecraft.
