Skip to main content

Towards Onboard Autonomy: Deployment of ML-based Anomaly Detection on Space-Qualified Hardware

Running

Running

Organisational Unit
Activity Type
11 April 2025

Duration: 6 months

Objective

The increasing complexity and criticality of space avionics systems necessitate advanced fault detection, diagnosis, and recovery techniques. This research investigates the deployment of autonomous health monitoring (AHM) strategies in space avionics to enhance system reliability, reduce mission downtime, and improve overall safety. The proposed AHM framework leverages data-driven techniques to analyze properly processed telemetries, enabling anomaly detection and prediction. To overcome the challenges posed by resource-constrained hardware, efficient algorithms and optimized data processing methods are explored. By integrating anomaly detection capabilities into the AHM system, early warning of potential failures can be achieved, leading to more proactive maintenance and reduced mission risks. This study aims to develop practical solutions for the integration of AHM into future space avionics systems, considering the unique requirements and constraints of the space environment.

Contract number
4000148070
Programme
OSIP Idea Id
I-2024-09563
Related OSIP Campaign
Visting Researcher Channel
Budget
6500€
Towards Onboard Autonomy: Deployment of ML-based Anomaly Detection on Space-Qualified Hardware