AI-Enabled Simulation Science for Manufacturing, Water Tech, and Agriculture: Fast, Accurate Modeling on Complex Geometry

Abstract:  

Modern engineering challenges—including advanced manufacturing, building design and operations, water technologies, and resilient agriculture—increasingly depend on our ability to simulate physics on complicated shapes: porous and architected materials, intricate flow passages, multi-material interfaces, and natural geometries such as plant canopies. Across these domains, three needs recur: (i) accurate representation of complex geometry, (ii) fast iteration for design and what-if exploration, and (iii) high-throughput generation of “AI-ready” datasets that span operating conditions, geometries, and uncertainty.

In this talk, I will describe an AI-enabled simulation-science pipeline aimed at meeting these needs. The core idea is to decouple fidelity from meshing bottlenecks by combining geometry-aware discretizations (e.g., adaptive octree methods) with immersed/embedded boundary approaches that accurately impose boundary conditions without repeatedly rebuilding body-fitted meshes. This enables rapid design iteration while retaining the reliability required for engineering decisions. I will then discuss how modern AI complements this workflow in two practical ways: (1) as a compact, queryable representation of geometry, and (2) as a driver for high-throughput parameter sweeps that produce large, structured datasets for training surrogates, enabling optimization, and quantifying uncertainty. The unifying message is that coupling robust simulation with AI is less about replacing physics, and more about turning physics into a faster, scalable engine for design, discovery, and decision support, particularly for electrochemical and electrokinetic water systems, building simulation and operations, manufacturing, and agriculture.

This is collaborative work with A. Krishnamurthy, M-C. Hsu, R. Anand, U. Passe (Iowa State), G. Scovazzi (Duke), A. Mani (Stanford) and H. Sundar (Tufts), and is funded by NSF, USDA-NIFA, and DoD.

Bio:

 Dr. Baskar Ganapathysubramanian is the Anderlik Professor of Engineering at Iowa State University. He received a B.Tech. from IIT Madras and a Ph.D. from Cornell University. He directs a curiosity-driven computational group focused on scientific computing, applied mathematics, and machine learning with applications to food, energy, and healthcare systems (me.iastate.edu/bglab). He is Director of the NSF/USDA-funded AI Institute for Resilient Agriculture (aiira.iastate.edu) and Associate Director of the Translational AI Center at Iowa State University.