.png){kind=logo}
NASA accumulated decades of information from satellites, spacecraft, Earth observations, simulations and scientific missions. The challenge is no longer simply collecting that information, but finding useful connections across vast amounts of data stored in different systems.
The agency is increasingly using AI to help researchers search, analyse and connect those datasets. NASA’s senior advisor on AI, Anna Steers-Smith, said the agency is using the technology as a tool for exploration and discovery rather than treating AI as an objective in itself. The push is relevant as NASA prepares for increasingly complex missions, where the amount of information generated is expected to grow significantly.
NASA has already used AI tools in planning for the Artemis II mission, helping teams bring together information for safety recommendations and risk-mitigation scenarios. According to Troy LeBlanc, chief information officer at NASA’s Johnson Space Center, analysing information from different sources could traditionally take experts weeks. AI tools were able to carry out parts of that work in seconds.
The agency is also examining how AI could support Mission Control. At Johnson Space Center, NASA has developed a prototype AI agent for SPARTAN flight controllers, who monitor the International Space Station’s power and thermal-control systems. The proposed system would function as a support resource, helping controllers sift through large volumes of information while human flight controllers remain responsible for real-time decisions.
NASA is also working with Microsoft on tools designed to make complicated Earth science datasets easier to access. Earth Copilot and Hydrology Copilot allow users to interact with NASA data using natural-language questions. Hydrology Copilot, developed with NASA’s Hydrological Sciences Laboratory, can work with information such as precipitation and soil moisture.
This could help users investigate drought conditions, flood risks and other water-related issues without requiring specialist knowledge of NASA's datasets. AI is also being used in weather forecasting. NASA foundational models trained on decades of Earth science data and physical models can help identify and simulate weather patterns more quickly. Processing satellite information closer to where it is collected could also speed up responses to events such as floods.
Also Read: NASA and Google Build AI Medical Assistant for Deep-Space Missions
NASA’s AI efforts extend beyond current missions. The agency has also experimented with AI agents that can search its public database of lessons learned from decades of programmes and missions.
A NASA-wide data and AI hackathon attracted more than 500 employees, far exceeding the expected participation. Other projects included machine-learning tools for assessing forest-fire severity.
For NASA, the larger opportunity lies in connecting information that has accumulated across generations of missions. As future projects become more complex, including plans associated with a potential Moon Base, AI could help scientists and engineers make better use of information that would otherwise remain difficult to search, compare or analyze.
