Can we build a precursor multi-instrument lunar foundation model for surface feature detection—and enable downstream decision intelligence for resource acquisition?

In support of FDL Lunarlab and the Luxembourg Space Agency’s ongoing lunar exploration objectives, we propose developing a multi-instrument foundation model for the Moon, capable of revealing and quantifying metallic signatures associated with impact ejecta. This work would build on existing FDL research which developed the first map of lunar thermophysical anomalies using physically-interpretable neural networks and provides a powerful ‘validation case’ to test the efficacy of the Foundation Model.