The “Digital Twins Webinar 5: Geospatial AI and Digital Twins in Contemporary Infrastructure” chaired by Liang Zhao from emory University explored the integration of big data, Artificial Intelligence, and Machine Learning for actionable insights. Keynote speakers delve into GeoAI opportunities, HPC Image processing, and foundation models, with panelists discussing AI applications at ESIP and NOAA divisions. Song Gao from University of Wisconsin-Madison unveiled pioneering research focused on bolstering food supply network resilience through advanced AI and a Geo Knowledge Graph, offering crucial insights for managing disruptions such as pandemics and natural disasters. The presentation underscores the vital trade-off between network resilience and efficiency, applicable across domains like food transportation and information communication. Dave Page with Oakridge National Lab is a senior researcher in 3D reconstruction and part of the management team of the fastest computer. He endorses digital twins for real-time monitoring and predictive analytics, emphasizes AI and problem-specific solutions while acknowledging challenges with nonlinear coordinate systems in geospatial applications. Gengchen Mai from University of Georgia introduces CSP, a self-supervised method for geospatial data with limited labels, enhancing model performance. It emphasizes the need for versatile foundation models and representative datasets for cross-domain advancements.
Understanding Food Supply Network Resilience via Geospatial Knowledge Graph and AI, Song Gao, University of Wisconsin-Madison
HPC Image Processing and Digital Twins at Oakridge National Lab, Dave Page, Oakridge National Lab
Panelist: Foundational Models for Geospatial AI: Applicability, Uniqueness, and Autonomy, Gengchen Mai, University of Georgia