TO-Agents: A Multi-Agent AI Pipeline for Preference-Guided Topology Optimization

2026-05-20

AI Summary

Designers often struggle to effectively translate qualitative intentions—such as aesthetic style or manufacturability—into parameters for topology optimization. This work proposes the first end-to-end topology optimization framework driven by natural language preferences, leveraging a multi-agent system to automatically parse design intent and generate corresponding optimization parameters. The approach introduces a multi-perspective vision-language critic agent coupled with a historical feedback mechanism, establishing a vision-language closed loop that iteratively refines 3D structures. Evaluated on two long-horizon design tasks, the method achieves a 60% alignment rate with human preferences and demonstrates up to a sixfold increase in success rate compared to a no-feedback baseline. Both human evaluations and agent-based scoring confirm its capacity for effective exploration and error correction.

Abstract

Topology optimization can generate efficient structures, but designers often must manually translate qualitative intent, such as desired visual style, product experience, or manufacturability into solver settings that are not directly tied to those preferences. We present TO-Agents, a multi-agent AI framework that connects natural-language design intent with iterative topology optimization. The framework converts a human-provided problem description into validated solver inputs, runs a topology optimization solver, renders the resulting 3D topology, and uses multi-view vision-language reasoning with an independent judge agent to critique each result and revise solver parameters. We evaluate the framework on two long-horizon design tasks: a cantilever beam benchmark and a phone-stand product design. In both tasks, the designer specifies an aesthetic preference for hierarchically branched structures inspired by natural tree morphologies, and the system performs four revision cycles across ten independent replicates. TO-Agents produces at least one preference-aligned design in 60% of trials for each case study, corresponding to up to 6x more successful trials than an ablated pipeline without visual or historical feedback. Judge scores and human evaluations show that the pipeline can identify effective parameter levers, recover from poor revisions, and expand design exploration. A manufacturing agent further post-processes top-ranked designs for additive manufacturing, enabling end-to-end intent-to-prototype design. We also identify failure modes, including overshooting, selective memory, misplaced tools, and incorrect parameter reasoning. These results suggest that agentic topology optimization can shift designers from low-level parameter tuning toward higher-level specification of form and function, while highlighting safeguards needed for reliable autonomous engineering design.

Problem

• topology optimization
• design intent
• preference-guided design
• multi-agent AI
• natural-language specification

Innovation

• multi-agent AI
• topology optimization
• preference-guided design
• vision-language reasoning
• additive manufacturing