Raytheon Intelligence & Space deployed a machine learning-based anomaly detection system for the US government's Joint Polar Satellite System to flag potential satellite hardware problems before they disrupt missions. The system was developed in partnership with C3.ai and reported publicly in 2021.

Working with C3.ai, Raytheon's Intelligence & Space division created a predictive anomaly detection capability embedded in satellite ground station software. The system takes real-time telemetry data from low Earth-orbiting satellites as input and produces early warnings of potential battery system failures or other anomalies as output, allowing ground operators to intervene before mission disruptions occur. The algorithm was trained on nearly 3.7 billion records of historical telemetry data spanning four years. At the time of the 2021 announcement, this was described as a pilot capability.