Aerospace, Google apply AI to satellite anomaly detection for LEO constellations

Aerospace and Google are developing an AI tool to detect satellite anomalies, streamline telemetry analysis, and support operations across large LEO constellations.
April 29, 2026
2 min read

Key Highlights

  • The project uses machine learning to monitor satellite behavior and detect anomalies more effectively than traditional threshold-based systems.
  • Unified telemetry interfaces allow operators to correlate signals with recent events, providing richer context for decision-making.
  • Automating data analysis aims to reduce response times and improve situational awareness in large satellite constellations.

CHANTILLY, Va. – The Aerospace Corporation in Virginia and Google Public Sector are working together to apply artificial intelligence to satellite operations. The effort centers on a proof-of-concept tool designed to help operators manage the growing complexity of proliferated low Earth orbit (pLEO) systems. 

As constellations expand from dozens of satellites to hundreds or more, engineers must process increasing volumes of telemetry across spacecraft, payloads, and ground networks.

Today, that information often sits in separate systems. When an issue appears, operators must manually compare data streams to determine whether a signal reflects a real fault or a transient condition. That process can delay response time, especially in high-tempo situations.

The new tool uses machine learning to monitor satellite behavior across multiple data sources. Instead of relying on fixed thresholds, the system looks for patterns that fall outside expected operating conditions.

AI-assisted anomaly detection

Combining telemetry into a single interface allows the software to flag subtle changes that may not trigger conventional alarms. It can also connect those signals to recent events, such as payload activity and environmental conditions, to provide additional context for operators.

The goal is to reduce the time engineers spend sorting through data and allow them to focus on diagnosing and resolving issues.

Scaling operations for large constellations

Tools that automate data correlation and highlight relevant signals could help teams maintain situational awareness across distributed systems. The concept also points to a move toward more predictive operations, where potential issues are identified before they affect performance.

The companies plan to continue developing the system as they explore how AI can support day-to-day satellite processes at scale. 

“Collaborations like this are an important part of Aerospace’s ongoing work to build connections with industry and accelerate the adoption of innovative technologies to support national priorities in space,” said Kevin Bell, senior vice president of Aerospace’s Engineering and Technology Group.

 

About the Author

Sign up for our eNewsletters
Get the latest news and updates

Voice Your Opinion!

To join the conversation, and become an exclusive member of Military Aerospace, create an account today!