Advancing Autonomy in Distributed Space Systems: Results From the Distributed Spacecraft Autonomy Experiment on Starling 1.0

The Distributed Spacecraft Autonomy (DSA) team at NASA’s Ames Research Center presents results of the DSA experiment onboard the Starling 1.0 mission. The DSA system showcases collaborative resource allocation for multi-point science data collection with four small spacecraft, highlighting autonomy in decision-making as a crucial factor for multi-spacecraft missions. Central to DSA’s operations is an autonomous GPS channel selection process used to optimize channel selection across the spacecraft swarm; the goal is to demonstrate fast, autonomous reactions to relevant scientific data. This is showcased by demonstrating the ability to capture ionospheric phenomena such as the Equatorial Ionization Anomaly and Polar Patches by utilizing relative Total Electron Content (TEC) measurements as sensor inputs. The DSA flight software architecture consists of three apps written in NASA’s Core Flight System framework: The Comm App, TEC App, and Autonomy (AUTO) App. The Comm App enables message routing over the Ad-Hoc Network of Starling 1.0. The TEC App processes GPS receiver data from the spacecraft bus and produces inputs to the Autonomy App based on the spacecraft geometry and ionospheric features extracted from the TEC signal. The Autonomy App fuses the rewards from the local TEC App with the rewards from the other spacecraft and uses a Mixed Integer Linear Program (MILP) solver to autonomously select the best set of observations to optimize performance while sending minimal information to other agents in the swarm. The DSA system’s autonomous reconfiguration ability is demonstrated, emphasizing its adaptability to natural phenomena without significant or direct tasking. The experiment’s success demonstrates the potential for autonomous systems in future space missions, enabling spacecraft to operate independently and efficiently.