IBM & NASA Launch AI Model to Map the Moon’s Hidden Features

IBM and NASA on Thursday unveiled an open-source AI model aimed at helping scientists analyze decades of lunar observation data and advance plans for a long-term human presence on the Moon.

The NASA-IBM Lunar Foundation Model, publicly available tool was developed specifically for lunar research. It was trained using more than 30 layers of data gathered by nine instruments across four NASA missions, including the Lunar Reconnaissance Orbiter.

The model is part of IBM and NASA’s Prithvi series of open foundation models, which are designed for applications ranging from geospatial analysis to weather research.

The AI can assist scientists in locating possible ice deposits in the Moon’s permanently shadowed areas, mapping craters to identify safer landing locations, and analyzing volcanic formations. These tasks have traditionally involved manually reviewing large numbers of maps and images or using less advanced machine-learning systems.

In benchmark testing, NASA and IBM said the model was up to 23% more accurate than commonly used methods at identifying important features on the lunar surface.

Scientists are particularly interested in lunar ice because it could provide access to water and oxygen , resources that may be crucial for establishing a future lunar base and producing rocket fuel for missions to Mars.

NASA’s Artemis program aims to return astronauts to the Moon in 2028, with the broader goal of testing technologies and systems needed for sustained lunar exploration and eventual missions to Mars.