Skip to Content
You are currently on the new version of our website. Access the old version .
BuildingsBuildings
  • Article
  • Open Access

3 April 2026

24 Pages

OM-GPT: A Knowledge-Augmented and Fine-Tuned Large Language Model for Prefabricated Building Operation and Maintenance Management

,
,
and
1
Department of Construction Management, School of Civil Engineering and Architecture, Shandong University of Science and Technology, No. 579 Qianwan Port Road, Huangdao District, Qingdao 266510, China
2
Department of Construction Management, School of Management, Guangzhou University, 230 Outer Ring, Rd., University Town, Panyu District, Guangzhou 510006, China
3
M.E. Rinker, Sr. School of Construction Management, University of Florida, Gainesville, FL 32611, USA
*
Author to whom correspondence should be addressed.

Abstract

The operation and maintenance (O&M) management of prefabricated buildings often struggles with fragmented knowledge and low reusability, relying predominantly on expert experience. While large language models (LLMs) offer a potential solution, their inherent hallucination issues significantly hinder practical application. To address these issues, this study proposes a knowledge base-augmented OM-GPT for prefabricated buildings O&M, built on a hybrid architecture that combines domain-specific fine-tuning with graph-based retrieval-augmented generation (GraphRAG). Specifically, it first fine-tuned the LLM Qwen2.5 using specialized O&M data to enhance its understanding of O&M tasks. It then constructed a multi-relational knowledge graph within a GraphRAG framework to effectively mitigate model hallucinations. Experimental results demonstrate that the Fine-Tuned Model achieved excellent Recall-Oriented Understudy for Gisting Evaluation (ROUGE) scores, validating the success of domain adaptation. In a five-dimensional evaluation, knowledge base-augmented OM-GPT significantly outperformed both GPT-4 and DeepSeek. Furthermore, two-way ANOVA tests confirmed the model’s advantages generalize across all five evaluation dimensions.

Article Metrics

Citations

Article Access Statistics

Multiple requests from the same IP address are counted as one view.