I-Corps: Translation Potential of an Onboard Shadow Twin For Self-Healing CubeSat (Cube-shaped Satellite) Power Systems

Award ID: 2637849

This I-Corps project is based on the development of an onboard shadow twin technology that makes small satellite power systems self-healing. Power system failures are the leading cause of small satellite mission loss, and most small satellites fly today without any onboard way to detect system degradation before it becomes catastrophic, relying instead on ground monitoring that cannot always maintain contact. This technology embeds a lightweight digital replica of the satellite's power system directly onto its onboard computer. This allows the spacecraft to compare live performance against an expected baseline to catch warning signs in real time, without new hardware or a ground link. This technology may enable longer, more reliable, lower-cost missions supporting Earth observation, communications, defense, and science, while also lowering barriers for smaller developers. In additon, the technology also has potential for use in terrestrial applications like remote power grids and distributed energy systems facing similar size and connectivity constraints. This I-Corps project utilizes experiential learning coupled with first-hand investigation of the industry ecosystem to assess the translation potential of an embedded digital-twin fault-prediction system for CubeSat (Cube-shaped Satellite) satellite power systems. Currently, existing digital-twin methods are too computationally heavy for small satellite hardware. This technology is an onboard embedded "shadow twin" system that makes CubeSat power systems self-healing by running a lightweight digital twin directly on the satellite's flight processor. It is able to detect soft faults and predicts failures in real time without requiring additional hardware. It uses a model-compression pipeline that shrinks a physics-informed power-system digital twin into a compact model able to run on a flight-grade, resource-constrained processor, paired with a fault-detection layer trained on early degradation signatures that precede failures. This may enable autonomous, preventive actions that extend mission life, by targeting the leading cause of CubeSat mission loss, power system failures. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Program Director: Ruth Shuman