NASA tests a new artificial intelligence chatbot to interpret complex data.
NASA has partnered with Microsoft to create an AI-powered chatbot aimed at analyzing large volumes of scientific data related to Earth.
Over the years, NASA has amassed a vast amount of complex data about our planet through its Earth observation satellites, which have been used to track climate change, monitor wildfires, among other things. However, understanding and disseminating this information has posed a considerable challenge. Now, thanks to generative artificial intelligence, Earth Copilot has been developed—a chat tool powered by AI in collaboration with Microsoft—aimed at making the complexities of satellite imagery and geospatial data more accessible, opening the door to new discoveries.
While currently only a limited number of researchers can interpret this data, Earth Copilot aims to extend its reach to a much broader audience. Users will be able to pose questions in a conversational manner, similar to how other AI-based chat tools like ChatGPT are used. Microsoft suggests that inquiries could be of the type: “What was the impact of Hurricane Ian on Sanibel Island?” or “How did the COVID-19 pandemic affect air quality in the U.S.?” Earth Copilot will access relevant datasets to provide answers that are comprehensible to users.
NASA's Earth Science Data Systems Division is responsible for gathering a wide variety of data from sensors and space instruments. This information ranges from atmospheric conditions to changes in land cover and ocean temperatures. However, the magnitude and complexity of this data can be overwhelming. For many, extracting relevant information involves navigating technical interfaces and understanding data formats—skills that non-specialized users rarely possess. Artificial intelligence could simplify this process, reducing the time needed to obtain information to just a few seconds.
Currently, Earth Copilot is in a testing phase by NASA scientists and researchers to ensure its reliability and safety. The next step will be to integrate it into NASA's Visualization, Exploration, and Data Analysis (VEDA) platform, allowing users to search, discover, and analyze Earth science-related data more easily.
