Radiation-Hardened Startracker incorporating Machine Learning Neural Networks for ESPA Missions Above LEO (HODUR)

As the Space Development Agency’s (SDA) Tracking Layer and Transport Layer satellite demonstrations are successfully under way, Ragnarok Industries, Inc. (Ragnarok) proposes to develop a Radiation-Hardened Startracker that incorporates Machine Learning Neural Networks for EPSA-class satellites above LEO named after HODUR (Norse “God of Darkness” and survivor of Ragnarok), that sustains > 300 krads of TID and > 84 MeV-cm^2/mg SEE of radiation effects, and is a crucial navigational sensor of the Guidance, Navigation and Control (GN&C) system required for the upcoming SDA Navigation Layer and Battle Management (Command, Control, and Communications: C3) Layers’ 3-axis stabilized satellites. Combining the monochromatic 1024 × 1024 rad-hard image sensor with our rad-hard flight computer produces a state-of-the-art navigation sensor whose performance is expected to be a significantly better solution than predecessors using obsolete radiation-hardened (RH) image sensors such as STAR250, STAR1000, and HAS2 for reduced SWaP-C. The HODUR startracker features rad-hard Analog-to-Digital conversion, hardened registers, and custom special-order radiation-resistant optical glass. HODUR startracker development will leverage Ragnarok’s experience in Radiation-Hardened by Design development, by incorporating the simulation of the radiation environment much earlier into the design phase that permits our team to select the crucial radiation-hardened microchips.