Chiller expert system development by using commercial and open-source large language models

Results in Engineering 2026 Open access

Read full text
Authors: Yen-Tang Chen, Dasheng Lee, Chia-Chi Kuo, Shih-Lung Chao, Ping-Hei Chen, Mohammad Amani
DOI: 10.1016/j.rineng.2026.109461
Cited by: 1
References: 11

Abstract

Large language models (LLMs) have shown promise in engineering knowledge transfer, but their use in industrial expert systems is limited by hallucinations and unreliable factual grounding. This study investigates the application of a retrieval-refined framework (Re2G) to improve the reliability of LLM responses in industrial chiller maintenance. Four system configurations were tested: RAG and Re2G pipelines with Gemma 2 9B and GPT-4o. Evaluation employed benchmark datasets, automatic metrics, LLM-as-judge scoring, and human ratings from 83 participants (74.7% with

Fields of Study & Subjects

BIM and Construction Integration Intelligent Tutoring Systems and Adaptive Learning Topic Modeling

Publication details

Journal / source
Results in Engineering
Publisher
Elsevier BV
Volume
29
Pages
109461–109461
Publication date
6 February 2026
Language
EN