Lunar AI Model Unveiled by NASA and IBM

NASA and IBM have unveiled an open-source lunar AI model that could revolutionize how scientists analyze data related to the moon's geography. The NASA-IBM Lunar Foundation Model combines multimodal and multi-resolution observations, allowing for faster identification of lunar ice deposits, volcanic features, and craters.

The model is trained on decades' worth of data collected by various sensors and instruments, including those from NASA's Lunar Reconnaissance Orbiter and the GRAIL mission. The dataset includes tens of thousands of images and maps of the lunar surface.

According to Kevin Murphy, the chief science data officer at NASA, 'collecting data is only part of the job.' He believes that AI can help turn large-scale data into new discoveries. The model could also aid in identifying safer landing sites for spacecrafts by better understanding the moon's terrain.