STATSAT: Statistical Control of Satellite De-Cluttering

R-DEX Systems proposes a novel Artificial Intelligence and Machine Learning (AI/ML) approach to decluttering space-based signal feeds for object tracking. R-DEX will leverage its background in robust machine learning, signal processing, and image-based tracking and classification to develop a system of neural network-based decluttering. The system will control view parameters of the sensors to allow for objects to be tracked within non-cluttered bands with R-DEXÆs SDEnet and advance statistical tools allowing for tuning the tracking and decluttering to new and uncertain conditions. R-DEX will also leverage its expertise in edge computing to minimize and package these systems in a small, ruggedized package for space-based and field computations. The end goal of the Phase I will be proving the feasibility of the SDEnet based control system and building a prototype of the system. The goal of the STATSAT system will be a Phase II/III transition with commercialization strategies being explored as part of the wider goal of placing robust and trustworthy models in the hands of commercial actors. Approved for Public Release | 23-MDA-11635 (9 Nov 23)