CISL Seminar: Understanding the community’s need for AI-ready open environmental data

Seminar
Apr. 20, 2022

1:00 – 2:00 pm MDT

Virtual

Speaker: Douglas Rao, NCICS and CISESS

 

Abstract

Artificial intelligence (AI), particularly machine learning (ML), has shown great potential to accelerate scientific discovery and data-based application development in earth and environmental sciences. These AI/ML research and applications are driven by the enhancement of computing capabilities and rapidly growing open environmental data in public and commercial platforms. However, researchers and developers often spend a large amount of effort in accessing and preparing the data for specific use cases. To proliferate AI/ML development with open environmental data, we need to understand the community’s need for data for efficient AI/ML development and develop community standards to improve future data services. Inspired by this need, the Earth Science Information Partners (ESIP) Data Readiness Cluster, including members from federal agencies, academic institutions, and private industry, developed a community roadmap for defining AI-readiness for environmental data and implemented a community survey to understand users' needs for AI/ML development. In this presentation, I will present the road map developed by the Data Readiness Cluster and summarize the AI-ready data survey results as well as the draft AI-ready data checklist. The presentation will also discuss future community activities on implementing the lessons learned from the community survey to develop thematic AI-ready data collections through community collaborations.

Bio

Douglas Rao is a Research Scientist at the North Carolina Institute for Climate Studies and Cooperative Institute for Satellite and Earth System Studies in Asheville, NC. He supports the development of the emerging NOAA Center for Artificial Intelligence and is also the co-lead for the ESIP Data Readiness Cluster to advance AI-ready data efforts.

Recording

Contact

Please direct questions/comments about this page to: