- The European Space Agency (ESA) has selected a Starion-led consortium to apply artificial intelligence (AI) to the automation of satellite ground segment systems testing and validation processes.
- The 14-month project will focus on three use cases identified in ESA’s Artificial Intelligence for Automation (A2I) Roadmap.
- The AI-based solutions are being developed for use both at ESA’s European Space Operations Centre (ESOC) and in commercial environments.
Starion has been awarded a contract by ESA to develop AI-based solutions for the automation of testing and validation processes for satellite ground segment systems. The development focuses on use cases identified within ESA’s Artificial Intelligence for Automation (A2I) Roadmap, originally defined by ESA’s Directorate of Operations in 2021.
Following the A2I Roadmap, ESOC has been demonstrating the benefits and impact of AI by augmenting a range of activities including mission planning, operations preparations and monitoring of spacecraft health as well as recommending actions to spacecraft operators.
In this latest project, a consortium led by Starion will focus on using state-of-the-art AI capabilities to help create test cases, automate the generation of test reports and support deviation analysis in validation/regression testing outputs. The aim is not only to address specific challenges and inefficiencies in existing operational processes but also to focus on user needs and workflows to ensure usability, foster understanding and trust in AI, and encourage adoption.
The solutions will take a modular approach to facilitate upgrades and integration with the latest AI technologies and will be incorporated into existing operational processes in a manner that minimises disruption and eases the transition to the new approach.
Aldo Barbieri, Managing Director of Starion Deutschland, said: “We are immensely proud to have been selected to lead this important initiative in the delivery of intelligent automation solutions to enhance testing efficiency and mission assurance. This achievement reflects our team’s commitment to innovation and technical excellence. Together with our partners, we bring a strong combination of skills in AI automation and space mission operations that will transform how testing and validation are performed in future space systems. By applying AI technologies, we aim to enhance efficiency, reliability and scalability, enabling faster, smarter and more resilient systems for the next generation of space missions.
“We are honoured by the trust placed in us by ESA, and we look forward to contributing our expertise to this important initiative. This milestone reinforces our strategic vision to lead the evolution of next-generation intelligent systems that power the future of mission operations.”
James Eggleston, Leader of the AI team in ESOC, said: “AI can provide powerful solutions for many of the tasks and processes involved in the development and testing of mission operations and ground segment systems, but in principle for any complex system. We’ve seen that the greatest benefits are obtained when these solutions are developed with current users, processes and tools in mind. We are looking forward to Starion building on its experience of supporting the integration of AI into operational environments, including previous projects within the A2I Roadmap, to address these new use cases.”
The development will leverage existing AI tools, including the MLOps platform AInabler, which Starion implemented for ESA to support the development, deployment and management of AI applications.
Although the new AI-based solutions are being developed in the context of system testing and validation at ESOC, they will be covered by the ESA Community Licence and are intended to be usable in non-ESOC environments. Further benefits will come from these new tools being developed in a way that ensures they can form part of an overall framework addressing the end-to-end needs of system engineering, from user requirements to detailed implementation and system validation.
The project consortium, led by Starion Deutschland, includes Fondazione Bruno Kessler, FEV etamax, Airbus Defence and Space, and Spacecube.
Further information
ESA’s AInabler platform was established to introduce and facilitate the use of AI and ML in operations and to serve as a common environment for the development and operation of supporting applications for current and future missions.
AInabler is a data platform providing support for all standard development languages, tools and libraries. This enables its users to build, (re-)train and deploy ML models to support domains, such as: mission planning support and optimisation; automated operations; ground stations; operations for telerobotics, autonomous and crewed systems; and pre-mission data extraction. The platform supports fast prototyping and testing, allowing fast judgement of the feasibility and possible value of a potential model or application, and inspection and evaluation of data quality.
For more information visit ainabler.space-codev.org/
Main image © ESA/J. Mai
