CGI is supporting the European Space Agency’s (ESA) Φsat-2 mission, with its Insula platform as a key enabler for AI-driven advances in Earth observation. The mission represents an important step forward in the development of new spatial data analysis capabilities.
Developed by the CGI team in Italy, Insula processes and harmonizes large volumes of satellite data, enabling advanced analytics such as trend detection, anomaly monitoring and predictive modeling. For the Φsat-2 mission, the platform provides calibrated and geolocated data sets to scientific communities and AI application providers, contributing to the success of the mission and the generation of new use cases.
AI models that use satellite data
Open to the public since June 2025, Insula has allowed more than 3,000 users to access data from satellite missions and has provided more than 10,000 multispectral images. The platform helps developers train AI models using real satellite data, accelerating innovation in a wide variety of applications and reinforcing the value of the Φsat-2 mission to the scientific community.
“The Φsat-2 mission demonstrates new approaches in Earth observation techniques, the results of which are shared through Insula,” says Nino Pace, Vice President of Consulting Services for Italy at CGI. “This innovation represents an important step in the application of AI to Earth observation, helping to generate new knowledge and improve the operational efficiency of the mission.”
Φsat-2, one of ESA’s most advanced small satellites, explores how AI running directly on board can improve Earth observation capabilities. The mission enables independent real-time functions such as cloud detection and automatic discarding of unusable images, maritime vessel detection and classification, street map generation, AI-based image compression and reconstruction from the ground, environmental monitoring, and support for emergency response scenarios, including fire detection. The scope of the mission also opens up new possibilities for future space initiatives.
ESA Open Data Policy
“The innovative use of onboard processing and AI on Φsat-2 demonstrates the potential to provide timely, high-quality information to Earth observation stakeholders,” highlights Mirko Albani, ESA’s Φsat-2 mission manager. “ESA’s open data policy expands access to Φsat-2 multispectral data, encouraging broader participation and demonstrating how small satellites can provide valuable information in various applications arising from this mission.”
Platform helps developers train AI models using real satellite data
“Our collaboration with ESA in Italy is based on more than a decade of experience in promoting innovation in the field of Earth, covering its observation, mission operations and planetary monitoring,” says Alberto Anaya, General Director and Business Unit Leader of CGI for Spain and Italy. “Through the expertise of our teams, we support the delivery of advanced technologies for the scientific community and contribute to the future of space exploration and the success of each mission in which we participate.”
