IGLC.net
EXPORT DATE: 2 June 2025

@CONFERENCE{Sepúlveda2025,
author={Sepúlveda, Italo and Alarcón, Luis F. and Barkokebas, Beda and Ebensperger, Antonia },
editor={Seppänen, Olli  and Koskela, Lauri  and Murata , Koichi  },
title={Developing a Genai Methodology for Data Analysis in Industrialized Construction: a Lean View}, 
journal={Proceedings of the 33rd Annual Conference of the International Group for Lean Construction (IGLC 33)}, 
booktitle={Proceedings of the 33rd Annual Conference of the International Group for Lean Construction (IGLC 33)}, 
year={2025},
pages={965-975},
url={http://www.iglc.net/papers/details/2393},
doi={10.24928/2025/0231},
affiliation={PhD Candidate, Department Construction Engineering and Management, Pontificia Universidad Católica de Chile, Santiago, Chile. Professor, Faculty of Architecture, Construction and Environment, Universidad Autónoma de Chile, Santiago, Chile, ilsepulveda@uc.cl, orcid.org/0000-0002-6019-9344 ; Professor, Department of Construction Engineering and Management, Pontificia Universidad Católica de Chile, Santiago, Chile, lalarcon@uc.cl, orcid.org/0000-0002-9277-2272 ; Assistant Professor, Department of Construction Engineering and Management, Pontificia Universidad Católica de Chile, Santiago, Chile, bbarkokebas@uc.cl,  orcid.org/0000-0002-0054-1320 ; Student Researcher, Department of Construction Engineering and Management, Pontificia Universidad Católica de Chile, Santiago, Chile, antonia.ebensperger@uc.cl, orcid.org/0009-0005-0843-5580 },
abstract={The research explores the potential of Generative Artificial Intelligence (GenAI) in enhancing data analysis processes within industrialized construction projects. The central question investigates whether a methodology can be developed to integrate GenAI into research workflows for construction projects. Existing studies highlight the challenges and opportunities of AI adoption in the construction industry but lack practical frameworks for its application in research, underscoring the need for this study.
The study employs GenAI across three phases studies: analyzing standardized data from 13 projects to identify common patterns and best practices, processing 57 interview transcriptions from industry leaders to assess readiness for industrialized construction, and comparing manual versus AI-supported analysis using 39 projects from an online industrialized construction database. The findings reveal that GenAI significantly reduces data processing time, enabling researchers to focus on in-depth analysis. Key lessons include the importance of prompt design, the context of data inputs, and the trade-offs between generic and customized AI models.
Building on these insights, the study proposes a GenAI-based methodology aligned with Lean Construction principles. The methodology was evaluated through a Likert-scale survey with seven construction professionals, confirming its clarity, feasibility, and applicability across various construction contexts. },
author_keywords={Industrialized Construction, Generative AI, GenAI, Data Analytics, Lean Construction. },
address={Osaka and Kyoto, Japan },
issn={2789-0015 },
publisher={ },
language={English},
document_type={Conference Paper},
source={IGLC},
}