CUE Distinguished Seminar Series: Simulating the Physical and the Connected: Neural Models for Structures and Networks

Advances in machine learning are transforming how we model and design engineered systems, from industrial components to large-scale infrastructure networks. This seminar presents recent developments in physics-based neural networks and neural operators for fast, accurate physics-based simulation. The first part focuses on individual structures and introduces models that accelerate design computations and improve generalization across geometries and parameters. The second part focuses on graph neural network (GNN) models for infrastructure networks and their use in developing digital twins of transportation systems. These GNNs enable rapid computation of network metrics—such as shortest paths, connectivity, and dynamic traffic flows—and can incorporate physics constraints like travel demand influence and flow conservation for enhanced accuracy and consistency. Case studies on urban networks demonstrate how such models can support resilient and data-driven decision making for asset management, mobility, and emergency response.

Dr. Hadi Meidani is an Associate Professor in the Department of Civil and Environmental Engineering at the University of Illinois at Urbana-Champaign (UIUC). His research focuses on uncertainty quantification, scientific and physics-informed machine learning, and fast computational methods for infrastructure and engineering design. He earned his Ph.D. in Civil Engineering, his M.S. in Electrical Engineering, and his M.S. in Structural Engineering from the University of Southern California (USC). Prior to joining UIUC, he was a postdoctoral scholar in the Department of Aerospace and Mechanical Engineering at USC and in the Scientific Computing and Imaging Institute at the University of Utah. Dr. Meidani is the Chair of the Machine Learning Committee of the ASCE Engineering Mechanics Institute. At UIUC, he is the Founding Chair of the AI in CEE Task Force in his department. Dr. Meidani is the recipient of an NSF CAREER Award for his work on fast computational models for infrastructure networks.