A Satisficing Approach to Generative AI-Driven Design of Nanoparticle-Infused Structural Materials: Integrating Mixture-of-Experts for Efficient Optimization

This project aims to revolutionize the design of nanoparticle-infused structural materials by combining Generative AI with a ‘satisficing’ approach—prioritizing practical, manufacturable solutions over purely mathematical optimization. Instead of searching for a single perfect design, our method will generate a variety of high-performing, user-acceptable, and manufacturable structures using AI-driven models. By integrating a Mixture-of-Experts (MoE) framework, which leverages specialized sub-models for different performance criteria, the project will enhance design efficiency while reducing computational costs. The collaboration between Canadian and international researchers will strengthen expertise in AI-driven materials engineering, benefiting both institutions through knowledge exchange, student training, and advancements in next-generation manufacturing technologies.

Faculty Supervisor:

Abu Syed Kabir

Student:

Partner:

Government Model Engineering College;JK Lakshmipat University

Discipline:

Engineering

Sector:

Education

University:

Carleton University

Program: