Arete Neuromorphic Star Tracker

Star-Trackers are ubiquitous in space applications, providing accurate absolute attitude calibration to satellites and other space-based systems. However, most celestial scenes are extremely sparse, featuring several stars on a fixed black background, wasting power and data processing on images with few features. This inefficient data usage as well as motion blur associated with platform motion, limit typical star-trackers to operation under 5 degrees per second. Low operational speed can lead to lost-in-space failure modes, limits application to highly dynamic platforms, and requires intermittent re-calibration for some platforms which reduces up-time. Event cameras are an emerging technology that excel at motion-blur free imaging of highly dynamic, sparse scenes that Areté believes is an enabling technology for next generation star-trackers. Areté will establish the viability of event cameras for star-tracking applications with current generation state-of-the-art event cameras and will design a prototype system known as the Areté Neuromorphic Star Tracker (AN-ST). The AN-ST will be based on existing Areté algorithms for event camera processing and star catalog mapping algorithms developed for previous programs. The AN-ST will build on Areté’s experience in accurate object tracking with event camera data, and existing codebases for stellar navigation, allowing an efficient trade study focused on novel components, and not recreation of existing algorithms/techniques. In Phase I, Areté demonstrated feasibility of using event camera data for star-tracking using data from a slewing ground-based detector. Phase-I results included demonstration of detection down to magnitude 8.1 stars with a suitable lens, tracking accuracy at the sub-pixel level, and processing of raw events into angular rates (as an alternative to a gyroscope). Our Phase II effort will focus on improving the performance of the algorithms developed in Phase I, as well as launching the first Generation 4 event camera into Low Earth Orbit (LEO) on a commercial bus. Orbital operation will provide test data in a mission-relevant environment, allowing us to further optimize the algorithmic chain for a space-based platform, as well as providing demonstrated flight heritage for event cameras which will be a crucial step in commercializing AN-ST