NASA and IBM’s Prithvi AI Model Reaches Orbit in First-Ever Deployment
With Prithvi, foundation models will cut satellite data delays. Researchers chose Prithvi for its strong generalization across Earth observation tasks and its open-source nature.
Topics
News
- NASA and IBM’s Prithvi AI Model Reaches Orbit in First-Ever Deployment
- AI Dispatch | 1–7 May 2026
- Saudi Aramco Bets $372.5m on Supercomputing for Hydrocarbon Discovery
- AI Reshaping Ranking and Value: Alphabet Closes In on Nvidia, Samsung Hits $1T
- The UAE’s Next Industrial Phase Takes Shape at MIITE 2026
- Trump Administration Eyes Regulatory Shift on AI as Cyber Threats Intensify
Research teams from Adelaide University and the SmartSat Cooperative Research Center in South Australia have successfully deployed NASA and IBM’s open-source Prithvi Geospatial AI foundation model in orbit — the first of its kind.
Trained on 13 years’ worth of Harmonized Landsat and Sentinel-2 datasets, Prithvi is equipped to facilitate a wide variety of Earth observation tasks, including flood mapping, cloud detection, disaster monitoring, and crop yield prediction.
A compressed version of Prithvi was uploaded to the South Australian government’s Kanyini satellite and to the Thales Alenia Space IMAGIN-e (ISS Mounted Accessible Global Imaging Nod-e) payload aboard the International Space Station.
Researchers chose Prithvi for its strong generalization across Earth observation tasks and its open-source nature.
“If Prithvi weren’t open source, I would have to train my own foundation model,” said Dr. Andrew Du, the project’s lead researcher, who is a postdoctoral researcher at Adelaide University and an AI engineer at the SmartSat Cooperative Research Center. “Having that model openly available saved a lot of time and effort.”
Kevin Murphy, chief science data officer at NASA Headquarters in Washington, whose office led the collaboration that created Prithvi, says its open-source nature was key for deployment. “By sharing these tools with anyone who wants to use them, we accelerate scientific and technological development into the future.”
Earth-observing satellites collect massive amounts of data on our planet. Processing it in orbit speeds up insights for researchers, but bandwidth constraints block big software updates, limiting satellites to lightweight, specialized AI models.
Foundation models, such as Prithvi, address this by handling multiple Earth observation tasks within a single setup. This milestone is an early demonstration of how such models could transform Earth observation.
“In the future, this could allow operators to interact with satellites in natural language, asking questions about onboard data or system status and receiving responses in a conversational way,” Du said.
The team is progressing its work in foundation models for planetary science, astrophysics, and biological and physical sciences.
