Slingshot Aerospace, in collaboration with DARPA, has developed an AI system called Agatha that can detect anomalies in satellite behavior and identify potentially harmful satellites within large satellite constellations, providing insights for space domain awareness.
Agatha has been trained on six decades of simulated constellation data and uses inverse reinforcement learning to evaluate objects' behavior and intentions in orbit, allowing it to detect small differences in satellite behavior that could indicate potential threats.
Slingshot is in discussions with U.S. government and commercial organizations about using Agatha as part of their space domain awareness services, as the increasing traffic in space makes it difficult for humans to sift through data and identify potential threats.
Axios reports that Agatha's data-agnostic nature allows it to be applied to other data types beyond satellite behavior, making it a powerful tool for monitoring space activities and maintaining safety and security.
Sources: TechCrunch, SpaceWar, DefenseScoop, NewsBytesApp, PayloadSpace, C4ISRNET, BusinessWire, SpaceNews, Axios.
This article was written in collaboration with Generative AI news company Alchemiq.
