Open-Source AI Model Released for Lunar Science

NASA and IBM have released an open-source AI model for lunar science, called the NASA-IBM Lunar Foundation Model. The model was trained on 17 years of observations from the Lunar Reconnaissance Orbiter and can be adapted to estimate ice prospectivity, map craters, track surface changes, and study irregular mare patches.

The model does not directly detect lunar ice, but rather estimates where it may be likely to occur or remain stable. This is an important distinction, as a value on the prospectivity scale is not direct evidence of frozen water presence.

The model's main training source is the Lunar Reconnaissance Orbiter, which has produced observations of the Moon across 17 years. The model's tasks are tied to scientific mapping rather than general-purpose chatbot behavior.