IBM, NASA release AI model for lunar mapping

Key facts

• IBM and NASA released the NASA-IBM Lunar Foundation Model.
• The model is open-source and trained on over 30 layers of data from four NASA missions.
• It can identify potential ice deposits and map craters on the Moon.
• Benchmark tests showed the model was up to 23% more accurate than existing methods.
• Lunar ice is crucial for future Moon bases and rocket fuel production.
• NASA's Artemis program plans to return astronauts to the Moon in 2028.

IBM and NASA on Thursday released an open-source AI model designed to help scientists analyze decades of lunar observation data and support plans for a sustained human presence on the Moon. The NASA-IBM Lunar Foundation Model is a publicly available AI tool trained on more than 30 layers of data collected by nine instruments on four NASA missions, including the Lunar Reconnaissance Orbiter.

The model joins IBM and NASA's Prithvi family of open foundation models, which span geospatial, weather, and other applications. It can assist researchers in identifying potential ice deposits in the Moon's permanently shadowed regions, mapping craters to select safe landing sites, and studying volcanic features. These tasks traditionally required manual analysis of maps and images or the use of lower-resolution machine-learning tools.

In benchmark tests, the model identified key features on the lunar surface up to 23% more accurately than widely used methods, according to NASA and IBM. Lunar ice is of particular interest to space agencies as it indicates the presence of water and oxygen, resources considered essential for a future Moon base and for producing rocket fuel for missions to Mars. NASA's Artemis program plans to return astronauts to the Moon in 2028, testing new technology for a sustained lunar presence and future Mars missions.