CEEM Seminar | Nick Vlassis | Rutgers University
Automating Engineering Mechanics Tasks with Generative AI and Foundation Models
Engineering mechanics increasingly relies on complex computational workflows for structural design, material characterization, and large-scale simulation studies, yet many of these pipelines still require extensive manual setup, expert intervention, and repeated numerical experiments. This seminar discusses how generative AI and foundation models can help automate parts of these workflows while preserving the underlying mechanics context. In structural design, we often seek undeformed configurations whose geometry and material distribution produce specific mechanical responses under loading, such as achieving prescribed deformation paths or satisfying geometric and functional constraints in systems ranging from inflatable biomedical devices, deployable aerospace structures, and soft robotics. Instead of exploring this design space through repeated forward simulations, latent-space generative models can learn relationships between structural configurations and target responses, enabling rapid generation of candidate designs that satisfy prescribed performance criteria. In geomaterials, the highly irregular microstructure of materials such as clays makes extracting particle-scale geometric information from microscopy data difficult, time-consuming, and often dependent on manual processing; foundation vision models provide a pathway to automate this step by identifying and generalizing particle geometries across complex datasets with minimal supervision, enabling scalable pipelines for geomechanics simulation analysis. As computational studies grow in size and complexity, interacting with these pipelines and extracting insights from their outputs becomes increasingly challenging. We demonstrate how AI assistants and agents can help researchers navigate these workflows, enabling higher-level interaction with simulation and design pipelines via natural language commands.
Nick N. Vlassis
Nick N. Vlassis is an Assistant Professor in the Department of Mechanical & Aerospace Engineering at Rutgers University. He received his diploma in civil engineering, with a specialization in structural design and analysis, from the National Technical University of Athens, Greece, in 2017. He earned his PhD from the Department of Civil Engineering and Engineering Mechanics at Columbia University in 2021. After graduation, he worked as a postdoctoral research scientist at Columbia University. He is the principal investigator of the Mech[Ai] group, which integrates computational solid mechanics with interpretable machine learning, graph learning, and generative AI to automate complex engineering workflows, with a focus on multiscale material modeling and design.