Craft Prospect launches ESA-funded AI4Ops project to advance onboard autonomy for small satellite missions

Glasgow-based space engineering company Craft Prospect Ltd (CPL) has partnered with GMV to deliver an advanced avionics demonstrator incorporating cutting-edge machine learning (ML) techniques to enhance onboard autonomy for small satellite missions.

Funded by the European Space Agency (ESA), AI4Ops brings together expertise in embedded AI and space systems engineering to develop autonomous operational capabilities targeting mission scenarios where onboard decision-making can deliver the greatest operational benefit. The work is applicable across a diverse range of institutional and commercial missions, including Earth observation, navigation, scientific, and lunar exploration applications.

Enabling autonomous operations in orbit

Two core technical applications have been identified as key enablers for autonomous spacecraft operations and will be demonstrated as part of the project:

• Responsive mission planning and scheduling, enabling spacecraft to dynamically adapt activities based on changing mission conditions and priorities

• Fault Detection, Isolation and Recovery (FDIR), allowing onboard systems to identify anomalies and respond autonomously to maintain mission performance and safety

Together, these technologies aim to reduce operational burden on ground teams while improving responsiveness to in-orbit events.

Demonstrating autonomy on a real mission

The developments are being designed against the requirements of a real flight scenario: ESA’s OPS-SAT VOLT (Versatile Optical Laboratory for Telecommunications) mission, led by Craft Prospect and planned for launch in 2026.

VOLT will host multiple advanced payloads spanning optical communications, quantum key distribution, and hyperspectral imaging. Supporting such a complex mission environment requires intelligent onboard systems capable of managing competing priorities while maintaining reliability and transparency. As a result, assurance and explainability of AI-enabled solutions form a central focus of the AI4Ops work.

The project has recently passed its System Requirements Review and has now entered the detailed design phase, progressing toward demonstration on flight hardware.

Words from the team

Lucy Donnell, Responsive Operations Lead at Craft Prospect, said:

“We look forward to presenting our results – the measured benefits of increased autonomy in VOLT, and more widely to other mission scenarios enabling the reduction of operational impacts and reaction times to events.”

Eleonora Mariotti, Avionics and Data Handling Engineer at ESA – ESTEC, commented: 

“As satellite constellations grow and orbital operations become more complex, the need for smarter, self-managing systems has never been more urgent. Our latest project is paving the way for on-board autonomous planning that is a strategic priority for ESA. By leveraging AI and ML, we are moving beyond traditional ground control to improve satellites to think and act independently. We are not just managing complexity, we are redefining what’s possible in orbit.”

Steven Kay, Robotics Engineer and Systems Engineer at GMV, said:

“At GMV, we’re committed to advancing onboard intelligence that makes future missions more agile, resilient, and autonomous. This collaboration allows us to demonstrate how reliable, explainable AI can unlock new levels of performance for future space missions.”

Advancing the future of autonomous space operations

As satellite missions grow more capable and complex, onboard autonomy is becoming a critical technology for improving efficiency, resilience, and scalability. Through AI4Ops, Craft Prospect and its partners aim to demonstrate how machine learning can safely and effectively enhance spacecraft operations, helping enable the next generation of intelligent space missions.