The lab’s mandate is four words long: accelerate the future of Earth observation. The tools include foundation models, onboard AI, and quantum computing.
Dr. Giuseppe Borghi has led ESA’s Φ-lab Division since June 2020, directing the agency’s open innovation laboratory at ESRIN in Frascati, Italy. His career spans nearly 30 years in the space industry, beginning with AI and robotics research at the Polytechnic University of Milan in the 1990s before transitioning to executive and managerial roles across the sector. He holds a PhD in robotics and artificial intelligence and a Master’s degree in executive general management.
Φ-lab operates as ESA’s bridge between emerging technology and operational Earth observation. The division’s current portfolio spans artificial intelligence, geospatial foundation models, quantum computing, onboard AI processing, and virtual reality applications for satellite data. In practice, that means building the tools that turn raw satellite imagery into actionable intelligence before it reaches the ground. The lab’s PhilEO geospatial foundation model, developed with e-GEOS and Leonardo Labs, was trained on global Copernicus Sentinel-2 data and represents ESA’s push toward large-scale AI analysis for Earth observation. In December 2025, the team demonstrated AI-generated near-real-time 3D cloud maps, a capability that improves the usability of optical satellite data by filling gaps caused by cloud cover.
The innovation network is expanding. In May 2025, SatNews reported on the inauguration of ESDI, the European Space Deep-Tech Innovation Centre at the Paul Scherrer Institute in Switzerland, extending Φ-lab’s reach into quantum technologies and advanced materials. In December 2025, SatNews covered AAC Clyde Space’s award from the ESA Φ-lab Sweden program to develop a next-generation onboard computer, a grant that illustrates how the lab seeds capability across European industry. The Φ-labNET initiative is establishing a network of national AI-for-EO hubs, with AI Sweden as its first international pilot.
At SmallSat Europe, Borghi brings the institutional perspective on where AI, EO, and smallsat architectures converge. For an industry debating onboard processing, edge computing, and autonomous satellite tasking, the question is whether Europe’s AI infrastructure can keep pace with the data its constellations generate.
The satellites produce the data. Φ-lab is building the intelligence layer that makes it useful.
