A Domain Adaptation of Large Language Models for Classifying Mechanical Assembly Components

2025-05-02

AI Summary

Structural data scarcity and subjective, inefficient annotation hinder early-stage design decision-making and behavioral modeling in functional classification of mechanical assembly components. Method: This paper proposes a domain-adaptive large language model (LLM) approach for functional semantic understanding. We introduce the first supervised fine-tuning of GPT-3.5 Turbo on the Oregon State Design Repository (OSDR), integrating functional semantic alignment with prompt engineering to automatically map natural-language functional descriptions to structured functional labels within the Function-Behavior-Structure (FBS) framework. Contribution/Results: Evaluated via zero-shot transfer on the ABC dataset, our method significantly improves classification accuracy while generating highly consistent, interpretable functional annotations. These high-quality labels robustly support behavioral modeling and conceptual design exploration, enabling more informed and scalable early-design reasoning.

Technology Category

• Cognitive Modeling & Cognitive Systems: Adaptive Behavior
• Machine Learning: Large Multimodal Models (LMMs)
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation

Abstract

The conceptual design phase represents a critical early stage in the product development process, where designers generate potential solutions that meet predefined design specifications based on functional requirements. Functional modeling, a foundational aspect of this phase, enables designers to reason about product functions before specific structural details are determined. A widely adopted approach to functional modeling is the Function-Behavior-Structure (FBS) framework, which supports the transformation of functional intent into behavioral and structural descriptions. However, the effectiveness of function-based design is often hindered by the lack of well-structured and comprehensive functional data. This scarcity can negatively impact early design decision-making and hinder the development of accurate behavioral models. Recent advances in Large Language Models (LLMs), such as those based on GPT architectures, offer a promising avenue to address this gap. LLMs have demonstrated significant capabilities in language understanding and natural language processing (NLP), making them suitable for automated classification tasks. This study proposes a novel LLM-based domain adaptation (DA) framework using fine-tuning for the automated classification of mechanical assembly parts' functions. By fine-tuning LLMs on domain-specific datasets, the traditionally manual and subjective process of function annotation can be improved in both accuracy and consistency. A case study demonstrates fine-tuning GPT-3.5 Turbo on data from the Oregon State Design Repository (OSDR), and evaluation on the A Big CAD (ABC) dataset shows that the domain-adapted LLM can generate high-quality functional data, enhancing the semantic representation of mechanical parts and supporting more effective design exploration in early-phase engineering.

Problem

• Classifying mechanical assembly components using LLMs
• Improving accuracy of functional data in early design
• Adapting GPT-3.5 Turbo for domain-specific function annotation

Innovation

• Fine-tuning GPT-3.5 Turbo for domain adaptation
• Automated classification of mechanical assembly parts
• Enhancing functional data accuracy via LLMs