AI tool detects unusual satellite behavior to improve space safety

A new artificial intelligence model has been developed to monitor satellites and identify unusual behavior, reducing the risk of collisions in Earth's orbit, according to recent research. This AI model, created by a team of international researchers and industry partners, is the first to predict anomalies and satellite motion by analyzing how light reflects off objects in space.

The project, part of the AI4 Space Safety and Sustainability program funded by the UK Space Agency, involves collaboration between universities and institutions across Australia, Canada, New Zealand, the United Kingdom, and the United States. Professor Massimiliano Vasile, director of the Aerospace Centre of Excellence at the University of Strathclyde and lead researcher, explained that the AI model can detect satellites behaving unexpectedly, predict their motion through space, and forecast future behavior.

Trained on massive amounts of satellite brightness readings gathered through telescopes, the model was further fine-tuned using simulations from the Aerospace Centre of Excellence and GMV. Once operational, the AI tool takes in real-time or recent light curves from ground-based observatories and flags any anomalies for human experts to investigate.

Testing showed the tool accurately identified unusual light curves 88% of the time and could distinguish between different satellite behaviors, such as spinning versus tumbling, which are crucial for satellite servicing and collision avoidance.

Vasile emphasized that this AI system demonstrates the potential for real-time anomaly detection, enabling human operators to quickly investigate flagged issues and take action to prevent collisions. The research team plans to explore the integration of additional data sources, such as radar, hyperspectral, and satellite orbit data, to enhance the tool's capabilities and provide more comprehensive insights for monitoring and safeguarding satellites.