MechE Colloquium with Faez Ahmed: Building Engineering-Native AI for Designing Physical Systems

Date: Friday, November 20, 2026 Location: 3-270 Time: 3:30pm-4:30pm ET Speaker: Faez Ahmed Affiliation: Associate Professor, MIT Mechanical Engineering Host: Professor Dan Frey Talk title: Building Engineering-Native AI for Designing Physical Systems Abstract: Generative AI has transformed how we create text, images, and code, yet it has not yet transformed how physical systems are designed. Engineering design is expressed through geometry, physics, and executable CAD and simulation workflows; it demands precision from small, expensive datasets; and it lacks the shared data and benchmarks that drove progress in language and vision. In this talk, I will present our work toward engineering-native AI: models, data, and tools built around these realities. I will begin with open datasets and benchmarks for engineering design across vehicles, ships, aircraft, and structures. I will discuss how datasets shape what models can learn, benchmarks define what counts as progress, and solvers determine whether a design can be trusted. I will then discuss domain-specific foundation models trained on large simulation-generated datasets, and how they generalize across problem settings in structural design. Next, I will turn to the more common setting in which data are scarce and evaluations are expensive. I will show how pretrained tabular foundation models can supply priors for multi-fidelity prediction and high-dimensional optimization. Finally, I will discuss AI that operates within engineers’ own tools, from generating editable CAD models to agents that learn to act in engineering workflows. I will conclude with a vision for AI that works alongside engineers across the full design process, from concept to manufacturable product. Bio: Faez Ahmed is the Esther and Harold E. Edgerton Associate Professor of Mechanical Engineering at MIT, where he leads the Design Computation and Digital Engineering (DeCoDE) Lab. His research lies at the intersection of artificial intelligence and engineering design, with a focus on generative AI, optimization, and datasets for designing physical systems. He received his Ph.D. in Mechanical Engineering from the University of Maryland and his B.Tech.–M.Tech. dual degree from IIT Kanpur, and was a postdoctoral fellow at Northwestern University before joining MIT in 2020. He has also worked in Australia’s rail and mining industry. His honors include the NSF CAREER Award, the AFOSR Young Investigator Award, the ASME Design Automation and Design Theory and Methodology Young Investigator Awards, a Google Research Scholar Award, and an Amazon Research Award. His teaching was recognized with the 2025 MIT MechE Keenan Award for Innovation in Undergraduate Education.