Launch of Open Source AI Model to Support Lunar Exploration

IBM and NASA launched an open-source AI model on Thursday designed to assist scientists in analyzing decades of lunar observation data. The model, trained on over 30 layers of data collected by 9 instruments across 4 NASA missions, can help identify potential ice deposits and map craters for safe landing sites. This model boasts a significant accuracy improvement of up to 23% in identifying key features on the lunar surface compared to widely used methods. It joins the Prithvi family of open-source AI models developed by IBM and NASA, which also includes applications in various fields.