IBM and NASA’s Marshall Space Flight Center today announced a collaboration to use IBM’s artificial intelligence (AI) technology to discover new insights from NASA’s vast amounts of earth and geospatial science data.of The collaboration will apply AI-based model technology to NASA’s Earth observation satellite data for the first time.
A Foundation model is a type of AI model that is trained on a wide range of unlabeled datasets, can be used for a variety of tasks, and can apply information about one situation to another. These models have rapidly advanced the field of natural language processing (NLP) technology over the past five years. IBM is pioneering applications for foundational models across languages.
Earth observations that allow scientists to study and monitor our planet are being collected at an unprecedented rate and volume. Extracting knowledge from these vast data resources requires new and innovative approaches. The goal of this work is to provide an easier way for researchers to analyze and derive insights from these large datasets. IBM’s underlying modeling technology has the potential to speed up the discovery and analysis of these data in order to rapidly advance our scientific understanding of our planet and our response to climate-related problems.
IBM and NASA will develop several new technologies to extract insights from Earth observations. One project trains an IBM Geospatial Intelligence Foundation model on NASA’s Harmonized Landsat Sentinel-2 (HLS) dataset, records of land cover and land-use change captured by satellites orbiting the earth. By analyzing petabyte-scale satellite data to identify changes in the geographic footprint of phenomena such as natural disasters, cyclical crop yields and wildlife habitats, this foundational modeling technology enables researchers to Helps provide critical analysis of environmental systems.
Another outcome of this collaboration is expected to be an easily searchable corpus of geoscience literature. IBM has developed NLP models trained on nearly 300,000 geoscience journal articles to organize the literature and facilitate the discovery of new knowledge. Fully trained models, including one of the largest AI workloads ever trained on Red Hat’s OpenShift software, use PrimeQA, IBM’s open source multilingual question answering system. In addition to providing resources for researchers, new language models for geosciences may be incorporated into NASA’s scientific data management and management processes.
“The beauty of the underlying model is that it can potentially be used for many downstream applications.” Said Rahul Ramachandran, Senior Research Scientist at NASA’s Marshall Space Flight Center Huntsville, Alabama. “Building these foundational models cannot be tackled by a small team,” he added. “We need teams that span different organizations to bring different perspectives, resources and skill sets.”
“The underlying model has proven successful in natural language processing, and now is the time to extend it to new domains and modalities that are important to business and society.” Said Raghu Ganti, Principal Scientist at IBM, said: “The application of foundational models to geospatial, event sequences, time series, and other non-linguistic elements within geoscientific data can provide highly valuable insights and information to a broader group of researchers, businesses, and citizens. Ultimately, it can empower more people to tackle some of our most pressing climate issues.
Other potential joint IBM and NASA projects under the contract include building foundational models for weather and climate prediction using the atmospheric observation dataset MERRA-2. This collaboration is part of NASA’s Open Source Science Initiative, an effort to build an inclusive, transparent, and collaborative open science community over the next decade.
For more information on this collaboration, IBM Research Blog.
All statements regarding IBM’s future direction and intent are subject to change or withdrawal without notice, and represent goals and objectives only.
sauce: https://www.ibm.com/in-en
