On-Board Anomaly Detection on a Flight-Ready System

Within the scope of an ESA funded activity, Airbus Defence and Space GmbH completed a research and development study in order to provide a novel dataset to ESA and develop a flight-ready system for on-board anomaly detection. This work includes the extraction of satellite telemetry data, the identification of anomalies, the development of machine learning models and the flight-ready system and finally the deployment of the machine learning algorithms via hardware acceleration. We present the benchmarking results of three accelerated ML algorithms from within the final flight-ready system.