IBM and the National Aeronautics and Space Administration (Nasa) have jointly launched an open-source AI model aimed at analyzing lunar data and bolstering plans for a long-term human presence on the Moon. The Nasa-IBM Lunar Foundation Model utilizes data from four Nasa missions, including the Lunar Reconnaissance Orbiter, encompassing over 30 data layers collected by nine instruments. This new tool allows researchers to identify potential ice deposits in the Moon’s shadowed regions, map craters for safe landings, and analyze volcanic features with increased efficiency. Benchmark tests demonstrate the model’s superior accuracy – up to 23 percent – compared to existing methods. The presence of lunar ice is crucial, indicating potential water and oxygen resources vital for a future Moon base and propellant production for Mars missions. This development supports Nasa’s Artemis program, targeting a return to the Moon in 2028 and preparation for future Mars exploration.