Knowledge Base Repository
In addition to research papers, the Design Society is developing several valuable resources for those interested in the study of design. These include a repository of PhD theses, a library of case studies and transcripts of design activities, and an archive of our newsletters. Please note that these resources are accessible exclusively to Design Society members.
Design Structure Matrix Modularization with Large Language Models
Jiang, Shuo; Luo, Jianxi
Type:
Year:
2026
Editor:
Langner, Christopher; Kreimeyer, Matthias; Panarotto, Massimo; Cascini, Gaetano; Browning, Tyson R.; Eppinger, Steven D.; Yassine, Ali A.
Author:
Series:
DSM
Institution:
Department of Systems Engineering, City University of Hong Kong, Hong Kong S.A.R. (China)
Page(s):
121-130
DOI number:
ISBN:
NA
ISSN:
NA
Abstract:
DSM modularization, the task of partitioning system elements into cohesive modules, is a fundamental combinatorial challenge in engineering design. Traditional methods treat modularization as a pure graph optimization, without access to the engineering context embedded in the system. Large Language Models (LLMs) offer a qualitatively different approach, combining structural reasoning with domain knowledge. In this paper, we apply LLM-based combinatorial optimization (LLM-CO) to DSM modularization. Our method achieves near-reference quality within 30 iterations without requiring specialized optimization code. Counterintuitively, domain knowledge, beneficial in sequencing, consistently impairs performance on more complex DSMs. We attribute this to semantic misalignment between the LLM's functional priors and the purely structural optimization objective, and propose the semantic-alignment hypothesis. Ablation studies identify the most effective input representation, objective formulation, and solution pool design for practical deployment. These findings offer practical guidance for deploying LLMs in engineering design optimization.
Keywords:
