On September 10, 2026, NASA and IBM released an open AI model that will help scientists analyze data about the Moon's surface.
The model, called the NASA-IBM Lunar Foundation Model, was trained on more than 30 layers of data from nine instruments across four NASA missions, including the Lunar Reconnaissance Orbiter.
It belongs to the family of open Prithvi models, which IBM and NASA already use for geospatial and weather tasks.
The tool can identify possible ice deposits in permanently shadowed regions, map craters to select safe landing sites, and study volcanic formations.
Previously, such tasks required manual analysis of maps and images or the use of less accurate machine learning methods.
In tests, the model showed accuracy 23% higher than that of widely used approaches, NASA and IBM reported.
Lunar ice is especially important: it indicates the presence of water and oxygen — resources needed for a future base and for producing fuel for flights to Mars.
The Artemis program plans to return astronauts to the Moon in 2028 to develop technologies for long-term presence and future missions to the Red Planet.
How will this model change the approach to planning lunar missions?
