The partnership aims to extract new insights from climate data, by combining IBM AI models with large amounts of NASA Earth observation data.
IBM has partnered with NASA to gain new insights on Earth’s climate through the power of AI technology.
The two organisations will use AI tech developed by IBM together with the large amounts of Earth observation and geospatial data that NASA has available to share.
Earth observation is the gathering of information about Earth’s physical, chemical and biological systems, usually through the use of satellite imaging.
IBM said the goal of this partnership is to provide an easier way for researchers to analyse and draw insights from these large data sets. The company plans to apply its foundation AI models – which are trained on broad sets of data – to speed up the analysis of this data.
The company said these types of AI systems have been used to advance natural language processing (NLP) technology in recent years. An example of an AI model that uses NLP is ChatGPT.
NASA senior researcher Rahul Ramachandran said these foundation models can be potentially used “for many downstream applications”.
“Building these foundation models cannot be tackled by small teams,” Ramachandran said. “You need teams across different organizations to bring their different perspectives, resources, and skill sets.”
The two organisations plan to work together on several projects to extract new insights from Earth observaton data.
For example, IBM plans to train a foundation model on NASA’s Harmonized Landset-Sentinel-2 dataset, which contains information about land cover and land use changes captured by satellites.
By analysing the satellite data, it is hoped this foundational model will identify changes in natural disasters, crop yields and wildlife habitats to help researchers analyse the Earth’s environmental systems.
IBM principal researcher Raghu Ganti said these AI models could lead to more people working on “our most pressing climate issues” by making new insights and information available to different groups.
“Foundation models have proven successful in natural language processing, and it’s time to expand that to new domains and modalities important for business and society,” Ganti said.
“Applying foundation models to geospatial, event-sequence, time-series, and other non-language factors within Earth science data could make enormously valuable insights and information suddenly available to a much wider group of researchers, businesses and citizens.”
10 things you need to know direct to your inbox every weekday. Sign up for the Daily Brief, Silicon Republic’s digest of essential sci-tech news.