Joint MechE Colloquium & Civil Engineering Seminar: Meta-neural Topology Optimization: knowledge infusion in Engineering design

Thumbnail

Event details

Date 13.05.2025
Hour 12:0013:00
Speaker Prof. Miguel Bessa, Brown School of Engineering, Brown University
Location Online
Category Conferences - Seminars
Event Language English
Abstract: Engineers learn from every design they create, building intuition that helps them quickly identify promising solutions for new problems. Unfortunately, topology optimization methods for Engineering design ignore past knowledge: every problem starts from a “blank canvas”. We propose a new method, meta-neural topology optimization, that infuses knowledge from past examples via a meta-learning strategy. The trained neural network creates effective designs that evolve towards optimal solutions faster than existing methods. Importantly, the proposed method does not suffer from “hallucination” issues that plague generative machine learning models, which would lead to catastrophic consequences in Engineering. Every design it creates is predicted with a simulation obeying Physics principles and leads to consistent designs, akin to conventional topology optimization methods. We believe that this fusion of machine learning and topology optimization will be a key enabler of future Engineering discoveries.

Biography: Professor Miguel Bessa and his research group envision a new era for the design of materials and structures using artificial intelligence. Professor Bessa received a PhD in Mechanical Engineering from Northwestern University in 2016 as a Fulbright scholar. After a short postdoctoral position at Caltech (2017) and a quick leap from Assistant to Associate Professor (2021) at Delft University of Technology, he joined the Solid Mechanics Group at Brown University in the Summer of 2022.

Practical information

  • General public
  • Free

Organizer

Tags

Joint MechE Colloquium & Civil Engineering Seminar: Meta-neural Topology Optimization: knowledge infusion in Engineering design

Share