IBM and NASA released an open-source AI model on September 10 to analyze decades of lunar data. The model maps craters, identifies potential ice deposits, and studies volcanic features, supporting upcoming crewed missions under NASA’s Artemis program and future human presence on the Moon.
Space exploration generates mountains of data, but raw observation is only half the battle. To turn decades of archives into actionable maps, IBM and NASA released an open-source AI model designed to help scientists examine the lunar surface. The tool arrives as space agencies ramp up preparation for returning astronauts to the Moon.
Inside the NASA-IBM Lunar Foundation Model
The newly unveiled artificial intelligence tool joins the Prithvi family of open foundation models developed jointly by IBM and NASA. These models span geospatial observation, weather forecasting, and other scientific applications. Training the lunar-specific model required digesting more than 30 layers of data collected by nine distinct instruments across four separate NASA missions, including information from the Lunar Reconnaissance Orbiter.
Traditionally, researchers had to manually sift through massive archives of maps and images or depend on lower-resolution machine-learning software to study the Moon. The new foundation model automates much of that heavy lifting. In benchmark testing, the system identified key surface features up to 23% more accurately than widely used methods.
Mapping Craters, Volcanism, and Shadowed Ice
- Identifying potential ice deposits hidden within the Moon’s permanently shadowed regions.
- Mapping surface craters to help mission planners select safe landing sites for future spacecraft.
- Studying volcanic features across the lunar terrain.
This capability is more than an academic exercise. Lunar ice serves as a crucial indicator for water and oxygen. Space agencies view these resources as essential for maintaining a permanent lunar base and for manufacturing rocket fuel destined for subsequent crewed missions to Mars.
The Open-Source Strategy Behind the Science
By making the model publicly available, the project builds directly upon a long-standing framework of shared exploration.
Artemis Timeline and the Road to Mars
The timing of the new AI release aligns closely with NASA’s broader deep-space roadmap. Under the Artemis program, the agency plans to return astronauts to the lunar surface in 2028. That upcoming mission will test critical technologies aimed at establishing a sustained human presence on the Moon and validating systems for future journeys to Mars.
