NASA and IBM develop new artificial intelligence model to map the lunar surface
NASA and IBM have prepared a new artificial intelligence model that will process decades of collected complex lunar data to map the surface in detail.
NASA and IBM scientists announced that they have created a new artificial intelligence model capable of processing scattered lunar data obtained from space missions over decades, aiming to map the surface down to the finest detail.
Lunar Data to Be Processed with Artificial Intelligence
The new artificial intelligence model developed in collaboration between NASA and IBM was designed to make sense of the satellite's complex and scattered data accumulated from the past.
Data from Past Missions Come Together
Years of robotic lunar missions had left behind a massive and fragmented pile of data that scientists had not been able to easily analyze until now.
This scattered information collected by spacecraft equipped with different sensors had been far from an accessible and common analysis method until now.
Benefiting from Image Recognition Models
Traditionally, such data was examined using limited transformer models such as SwinV2-B, which were developed to increase accuracy in photo recognition and related visual tasks.