Exploring the power of Large Language Models: Automated compliance checks in architecture engineering and construction industries

Liu, Xiaoyu 2025. Exploring the power of Large Language Models: Automated compliance checks in architecture engineering and construction industries. MPhil Thesis, Cardiff University. Item availability restricted.

Preview	PDF (Thesis) - Accepted Post-Print Version Download (3MB) | Preview
	PDF (Cardiff University Electronic Publication Form) - Supplemental Material Restricted to Repository staff only Download (286kB)

Abstract

In recent years, Large Language Models (LLMs) have emerged as one of the most rapidly evolving sectors within the field of artificial intelligence. These models have increasingly penetrated various industries, becoming integral to our professional and daily lives. This study focuses on the customization of six specialized LLMs, each injected with professional knowledge pertinent to the Architecture, Engineering, and Construction (AEC) domain. Initially, we leveraged a bespoke integrated model to investigate the feasibility of implementing an Automated Compliance Checking (ACC) process within this industry. Subsequent phases involved employing a datagenerating model for preparatory data tasks, followed by the deployment of three specialized models serving distinct target objectives. These models were interlinked to construct a prototype for the ACC process. Finally, a professional model dedicated to data analysis was utilized to quantitatively assess the performance of the entire ACC prototype in regulatory compliance checks. Drawing on the results of this analysis, we provide a comprehensive evaluation of the ACC prototype's effectiveness. This thesis elucidates the profound potential of LLMs to revolutionize compliance checking within the AEC sector, highlighting the intricacies of developing, implementing, and refining a specialized ACC process. Through the customization and application of LLMs, this research not only showcases the practical viability of ACC but also advances our understanding of LLMs' adaptability and utility in ii specialized domains. The findings contribute significantly to the ongoing discourse on the integration of artificial intelligence in industry-specific applications, offering valuable insights for future advancements and implementations.

Item Type:	Thesis (MPhil)
Status:	Unpublished
Schools:	Schools > Engineering
Uncontrolled Keywords:	1) . Automated Compliance Check 2) . Large Language Models 3) . AEC industry 4) . AI downstream applications 5) . Artificial Intelligence 6). Natural Language Processing
Date of First Compliant Deposit:	15 April 2025
Last Modified:	15 Apr 2025 15:09
URI:	https://orca.cardiff.ac.uk/id/eprint/177710

Actions (repository staff only)

Edit Item