NASA and IBM have released an open source artificial intelligence model designed to help scientists analyze lunar data and support plans for a long term human presence on the Moon.
The NASA IBM Lunar Foundation Model was trained using more than 30 layers of data from nine instruments aboard four NASA missions, including the Lunar Reconnaissance Orbiter.
The model can help researchers locate possible ice deposits in permanently shadowed areas, map craters for safer landing sites and identify volcanic features. NASA and IBM said it performed up to 23% better than widely used methods in benchmark tests.
Scientists consider lunar ice important because it could provide water and oxygen, resources that may support future lunar bases and the production of rocket fuel for missions to Mars.
The model is part of NASA and IBM’s Prithvi family of open foundation models covering areas such as geospatial and weather research.
NASA’s Artemis programme aims to return astronauts to the Moon in 2028 as part of efforts to develop technologies for sustained lunar exploration and future Mars missions.
Leave a comment