Wenqian Zhang, Yangyi Huang, Zigeng Xu, Weiyang Liu, Zhen Liu · 2026 · arXiv (Cornell University)
DOI: 10.48550/arxiv.2610.26861 · cited by 0 works (OpenAlex) · open-access PDF
Alors que les agents LLM peuvent générer des assemblages CAD visuellement plausibles, ces conceptions échouent souvent à remplir leurs fonctions physiques prévues. Nous étudions la conception de CAD fonctionnelle, un environnement dans lequel les agents interprètent les demandes sous-spécifiées pour construire et réviser de manière itérative des mécanismes qui satisfont aux exigences fonctionnelles dans la simulation physique. La vérification de ces exigences est non triviale: les mécanismes candidats différents peuvent réaliser la même fonction à travers des composants, des contrôles et des séquences opérationnelles complètement différentes. Pour combler ce fossé, nous proposons FunCAD-Harness, un cadre qui traduit les exigences abstraites en tests physiques spécifiques aux candidats et utilise
Machine translation (Meta M2M100); the English original follows.
While LLM agents can generate visually plausible CAD assemblies, these designs often fail to fulfill their intended physical functions. We study functional CAD design, a setting where agents interpret underspecified requests to construct and iteratively revise mechanisms that satisfy functional requirements in physical simulation. Verifying these requirements is nontrivial: different candidate mechanisms may realize the same function through entirely different components, controls, and operating sequences. To bridge this gap, we propose FunCAD-Harness, a framework that translates abstract requirements into candidate-specific physical tests and uses simulation feedback to guide design revisions. To systematically evaluate this capability, we introduce FunCADBench, a benchmark spanning 46 mechanism families with interdependent and multi-step operations. Experiments across frontier LLM agents show that FunCAD-Harness exposes subtle mechanical failures invisible to visual inspection, substantially improving functional success rates over standard simulation-augmented baselines. Ultimately, our work marks a step toward CAD agents capable of translating human intent into working mechanisms.
Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese
Topics: CAD, Computer science, Construct (python library), Function (biology), Mechanism (biology), Bridge (graph theory)
GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.26861, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txtSources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange