Path Planning and Machine Intelligence for Collaborative Inspection of Non-Cooperative Space Objects

The proposed solution will enable a Multi-Agent Satellite System to collaboratively perform inspection, identification, classification and characterization of non-cooperative resident space objects. The project will test a prototype algorithm for identification of features on a space object, classification and characterization in a spacecraft formation flight testbed, using realistic target object mockups and lighting conditions. Furthermore, the project will use computer simulations to demonstrate a guidance algorithm enabling multiple spacecraft to collaboratively inspect and map a target object safely and efficiently, with emphasis on collision avoidance, avoidance of surface glare and avoidance of direct exposure of cameras to sunlight. Therefore, the capabilities enabled by the proposed solution form the baseline for on-orbit servicing, in-space assembly, debris removal and space asset protection operations.  The deliverables of the proposed Phase I STTR project are: (1) demonstrated machine intelligence prototype for autonomous and collaborative feature identification leading to space object classification and characterization; (2) demonstrated autonomous and collaborative path planning for optimized space objects inspection; (3) feedback from major corporations in the field of on-orbit servicing and in-space assembly towards the commercialization of the solution.