NASA and IBM have jointly released the Lunar Foundation Model, an open-source artificial intelligence tool designed to enhance lunar scientific research. This AI model was trained using nearly two million tile bundles derived from 17 years of data collected by NASA’s Lunar Reconnaissance Orbiter.
According to The Decoder, the model significantly improves predictions related to lunar polar ice deposits, reducing errors by up to 22 percent compared to the strongest existing models. This advancement could provide more accurate insights into the Moon’s surface composition and resources.
For Japanese markets, the development highlights the growing role of AI and space data in driving innovation, potentially influencing technology sectors focused on space exploration and satellite data analysis.