Agentic Large Language Models for Conceptual Systems Engineering and Design
Journal of Mechanical Design · 01.05.2026 · 148 (5)
DOI: 10.1115/1.4070328 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Mechanical Design, online erschienen am 13.01.2026. Verfasst von Soheyl Massoudi und 1 weiteren Autor:innen. Die Arbeit wurde bisher 10-mal zitiert (Crossref) und nennt 55 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Early-stage engineering design involves complex, iterative reasoning, yet existing large language model (LLM) workflows struggle to maintain task continuity and generate executable models. We evaluate whether a structured multi-agent system (MAS) can more effectively manage requirements extraction, functional decomposition, and simulator code generation than a simpler two-agent system (2AS). The target application is a solar-powered water filtration system as described in a cahier des charges. We introduce the design-state graph (DSG), a JSON-serializable representation that bundles requirements, physical embodiments, and python-based physics models into graph nodes. A nine-role MAS iteratively builds and refines the DSG, while the 2AS collapses the process to a generator–reflector loop. Both systems run a total of 60 experiments (2 LLMs—Llama 3.3 70B versus reasoning-distilled deepseek R1 70B × 2 agent configurations × 3 temperatures × 5 seeds). We report a JSON validity, requirement coverage, embodiment presence, code compatibility, workflow completion, runtime, and graph size. Across all runs, both MAS and 2AS maintained perfect JSON integrity and embodiment tagging. Requirement coverage remained minimal (less than 20%). Code compatibility peaked at 100% under specific 2AS settings but averaged below 50% for MAS. Only the reasoning-distilled model reliably flagged workflow completion. Powered by deepseek R1 70B, the MAS generated more granular DSGs (average 5–6 nodes) whereas 2AS mode collapsed. Structured multi-agent orchestration enhanced design detail. Reasoning-distilled LLM improved completion rates, yet low requirements and fidelity gaps in coding persisted.
Zitieren
APA
Massoudi, S., & Fuge, M. (2026). Agentic Large Language Models for Conceptual Systems Engineering and Design. Journal of Mechanical Design, 148(5). https://doi.org/10.1115/1.4070328
Vancouver
Massoudi S, Fuge M. Agentic Large Language Models for Conceptual Systems Engineering and Design. Journal of Mechanical Design. 2026;148(5). doi:10.1115/1.4070328
BibTeX
@article{Massoudi2026,
author = {Massoudi, Soheyl and Fuge, Mark},
title = {Agentic Large Language Models for Conceptual Systems Engineering and Design},
journal = {Journal of Mechanical Design},
year = {2026},
volume = {148},
number = {5},
doi = {10.1115/1.4070328}
}
EUVIMED