“Artificial Intelligence (AI) will enrich traditional approaches to modeling complex systems rather than replace them.”— Google DeepMind
The future of Computer Aided Engineering (CAE) lies at the intersection of AI, traditional simulation, and scientific knowledge. Traditional methods alone are insufficient for increasingly complex Science and Engineering (S&E) problems. By scaling explainable predictive AI and combining it with domain specific generative models informed by scientific knowledge and classical simulation, a powerful multi agent AI paradigm can be established for next generation modeling and simulation.
This two-and-a-half-day short course introduces mechanistic computational intelligence tools and concepts, focusing on how hierarchical neural networks and agentic AI systems integrate with finite element analysis to improve accuracy, reduce computational cost, and automate workflows.
June 22–24, 2026
1801 Maple Avenue, Evanston, IL 60201
2.5 Days
In-Person or Remote (via Zoom)
Familiarity in Agentic AI for CAE and how it addresses fundamental challenges in Computational Science and Engineering.
Insights into tackling simulations that require unparalleled precision and accuracy or involve extremely large degrees of freedom via HIDENN-AI’s technology.
Understanding of how to seamlessly integrate data-driven AI training and data-free AI solving.
Hands-on experience with HIDENN-AI’s cutting-edge software applied to real-world industry problems through live demonstrations.
Light breakfast and introduction
Mechanistic Computational Intelligence for Science and Engineering Overview
Interpolation and Neural Networks: Constructing a Hierarchical Deep Learning Neural Network (HiDeNN)
Break
Extending HiDeNN to Convolution HiDeNN (C-HiDeNN)
Extending C-HiDeNN to C-HiDeNN Tensor Decomposition (C-HiDeNN-TD) — Scalability through Tensor Decomposition
C-HiDeNN-TD for training/learning
Lunch Break
HiDeNN interface with commercial program (Abaqus/Ansys) for Linear and Nonlinear Mechanics (repeated next day)
REPEAT: HiDeNN interface with commercial program (Abaqus/Ansys) for Linear and Nonlinear Mechanics
Light breakfast
Ex-HiDeNN Training and Solving
Bayesian Interpolation Neural Networks: learning simulation data with uncertainty
Break
C-FEM agent: automatic error mitigation with only one-step remeshing
Solving parametric PDEs with C-HiDeNN-TD
Sketch to 3D geometric representation for complex geometries
Lunch Break
Tutorial: C-FEM agent (repeated next day)
Half day
REPEAT: Tutorial: C-FEM agent
Light breakfast
Immersed Tensor Decomposition (ITD) for arbitrary geometries
GO-MELT for fast Additive Manufacturing (AM) simulation
Break
Outlook for Future/Impact of Mechanistic Computational Intelligence — One Engineering Software for the Entire Engineering Process
Co-founder
Northwestern University
PhD from Caltech, Walter P Murphy Professor of Mechanical Engineering at Northwestern University, author of "Mechanistic Data Science for STEM Education and Applications" and "Nonlinear Finite Elements for Continua and Structures." Ranked #25 in the world and #15 in the United States in the 2024 Ranking of Top 1000 Scientists in Mechanical and Aerospace Engineering by Research.com.
Co-founder
UT Dallas
PhD from Northwestern University, Professor and Associate Head of Mechanical Engineering at UT Dallas, Fellow of the American Society of Mechanical Engineers.
Software Developer
HIDENN-AI
PhD from Harvard University.
Software Developer
HIDENN-AI
PhD from Northwestern University.
Software Developer
HIDENN-AI
PhD from Northwestern University.
Software Developer
HIDENN-AI
PhD from Northwestern University.
| Format | Early (before Apr 30) | Regular (after Apr 30) | Early Academic | Regular Academic |
|---|---|---|---|---|
| In-Person | $1,725 | $1,950 | $950 | $1,150 |
| Remote (Zoom) | $1,450 | $1,750 | $750 | $950 |
Prices listed are per person (USD).