AI Model to Map Moon's Craters and Search for Lunar Ice
NASA and IBM have jointly released an AI model for Moon research, called the NASA-IBM Lunar Foundation Model. The pre-trained model is designed to study the Moon's surface, and can be adapted for various tasks such as mapping craters, searching for young volcanic formations, and assessing areas where ice may exist near the lunar poles.
The model was trained on a dataset of roughly 2 million patches of images of the lunar surface, gathered from various missions including the Lunar Reconnaissance Orbiter. It outperformed other foundational models in tests on all evaluated tasks.
One of the key applications of this AI model is to identify permanently shadowed regions where ice can remain stable for billions of years. This data will help scientists study the Moon's history and potentially identify resources for future space missions.