NASA and IBM Release AI Model to Accelerate Lunar Research

After decades of collecting vast amounts of information about Earth’s nearest neighbor, NASA and IBM have introduced a new artificial intelligence tool designed to make that data more accessible. Announced on Thursday, September 10, the NASA-IBM Lunar Foundation Model is an open-source resource now available at no cost on Hugging Face, aimed at helping scientists process the massive datasets gathered from lunar missions.

NASA’s chief science data officer and acting chief data and AI officer, Kevin Murphy, emphasized the importance of this advancement. “NASA has spent decades building an extraordinary scientific record of the moon, but collecting data is only part of the job,” Murphy said. He noted that the new model enables researchers to perform complex tasks such as mapping craters, identifying recent volcanic features, and modeling potential water ice deposits near the lunar poles.

The initiative highlights a growing trend in planetary science, where space agencies and private companies globally are turning to AI to boost productivity, lower costs, and streamline analytical workflows. This collaboration follows previous partnerships between NASA and IBM, which produced AI models for Earth observation and heliophysics.

The lunar model was primarily trained on data from the Lunar Reconnaissance Orbiter (LRO), NASA’s longest-serving mission to the moon. Over its 17-year operation, the orbiter has captured high-definition imagery that forms a nearly complete mosaic of the lunar surface. The volume of data collected by LRO alone surpasses that of all other NASA planetary missions combined.

Murphy stated that the project demonstrates the potential of applying AI to NASA’s petabytes of scientific information. “The NASA-IBM Lunar Foundation Model shows what’s possible when we bring AI to NASA’s petabytes of scientific data,” he said. “That’s a real opportunity we see with AI: turning large-scale data into new discoveries.”