SBIR/STTR Award attributes
Space missions continue to increase in number, complexity, and time amp; cost constraints. To lead technological advancements and successfully execute these missions, NASA desires new amp; robust onboard automated fault management technologies that address the full range of hardware amp; software faults, are transparent amp; reusable across platforms. This will lead to reduced costs and improved autonomy, resilience, amp; mission quality especially in missions that cannot afford comprehensive fault management and have a higher mission risk tolerance.nbsp;Global Technology Connection, Inc., proposes ARADISS (Adaptive Real-time Anomaly Detection and Identification for Space Systems) framework applicable to virtually all electrically powered systems. It involves physics-guided machine learning models to detect and simultaneously locate faults. The feature learning ML models continue to learn in real-time to adapts to gradual system degradations which avoid extensive model training requirements. This technique has demonstrated comprehensive fault coverage with a high detection rate and a low detection latency in extensive tests on automobiles. Meaningful physical correlations to battery voltage fluctuations make this approach extremely transparent and immediately transferable to other platforms. These algorithms are computationally inexpensive to run and can be implemented on-board small space missions.nbsp;In Phase I, we propose to validate this framework on a UAV and show feasibility demonstration and applicability to future space mission platforms like NISAR, SWOT, Dragonfly, and SPHEREx.In Phase II, our team will identify a space platform, tune, and test our algorithms in simulated environments for transition and space deployment in Phase III.Aggressive commercialization activities would be carried out throughout all phases of this program, and Phase III will concentrate on technology transition for NASA and other commercial space applications
