The world’s first geospatial AI foundation model to function on space-based systems is the open-source Prithvi model, which was successfully sent into orbit by a group of Australian researchers. 

Thirteen years of Earth observation data from NASA’s Landsat and the European Space Agency’s Sentinel-2 satellites were used to train the model, which was created in collaboration with IBM. 

The South Australian government’s Kanyini satellite and the International Space Station’s IMAGIN-e payload both received a condensed version of Prithvi. The model’s ability to identify clouds and floods in various orbital computing environments was evaluated by researchers. 

Prithvi eliminates the need to send unprocessed satellite imagery back to Earth for examination by processing geographical data while it is in orbit. According to experts, this might greatly speed up the response to natural disasters like landslides, wildfires, and floods. 

Additionally, the AI model can be modified for disaster management, environmental monitoring, and crop prediction. According to researchers, using AI directly in orbit could enhance global Earth monitoring capabilities and speed up real-time decision-making. 

Source – News9 Live