IBM and NASA team up to open source geospatial AI models to advance climate science research

According to foreign media Fagen Wasanni Technologies, IBM and NASA announced that they will cooperate to use IBM’s watsonx.ai and NASA’s satellite data to build an open source geospatial AI basic model on the open source AI platform Hugging Face. This model will become the largest geospatial basic model on Hugging Face and the first open source AI basic model built in cooperation with NASA

▲ The picture comes from the AI ​​tool page of watsonx.ai

Because environmental conditions change daily in climate science, obtaining up-to-date data remains a huge challenge. According to NASA estimates, scientists will have 250,000 terabytes of data from new missions by 2024. However, scientists and researchers still face several obstacles in analyzing these large data sets. Therefore, IBM and NASA believe that by harnessing the power of artificial intelligence, the situation can be improved

IBM began building a basic artificial intelligence model called watsonx.ai earlier this year for processing geospatial data as part of the Space Act Agreement signed with NASA

▲ The picture comes from the AI ​​tool page of watsonx.ai

IBM and NASA collaborated for a year to train this model using Harmonized Landsat Sentinel-2 satellite data. Researchers are continuously improving the model, and as of now, its efficiency has increased by 15% compared to its initial release.

In addition, officials stated that the model has been fine-tuned to a certain extent on the annotated data, and the basic model can be used for tasks such as monitoring deforestation, predicting crop yields, and detecting and monitoring greenhouse gases. At the same time, researchers from IBM and NASA collaborated with Clark University to add scientific applications to the model, such as "time series segmentation" and "similarity study"

According to reports, IBM plans to offer a commercial version of the watsonx.ai geospatial model through the IBM Environmental Intelligence Suite (EIS) later this year. In the meantime, an open source version of the model is already available on Hugging Face, where those interested can access it

Sriram Raghavan, Vice President of IBM Research AI, states, "The crucial role of open-source technology in accelerating discoveries in critical areas such as climate change has never been more evident."

IBM claims that by combining its reusable AI systems with NASA’s Earth satellite data warehouse, humans can harness the power of collaboration to implement faster, more impactful solutions to improve the planet

Jeff Boudier, the Product and Growth Lead at Hugging Face, stated, "AI remains a science-driven field, and science can only progress through information sharing and collaboration. That's why open-source AI and model datasets are crucial for the sustainable development of AI, as they ensure that the technology can benefit as many people as possible."

Kevin Murphy, NASA's Chief Science Data Officer, said: "We believe that basic models have the ability to change the method of analyzing observational data, thereby helping us understand the Earth more deeply and jointly assist the development of mankind."

The above is the detailed content of IBM and NASA team up to open source geospatial AI models to advance climate science research. For more information, please follow other related articles on the PHP Chinese website!