TY  - CONF
TI  - Establish a lean AI-driven defect management framework
C1  - Singapore, Singapore
C3  - Proceedings of the 34th Annual Conference of the International Group for Lean Construction (IGLC 34)
SP  - 120
EP  - 131
PY  - 2026
DO  - 10.24928/2026/0204
AU  - Tolulope, Akande
AU  - Shang, Gao
AU  - Oraee, Mehran
AD  - Doctorate Candidate, Faculty of Architecture, Building and Planning, The University of Melbourne, Melbourne, Australia, akandetc@student.unimelb.edu.au, orcid.org/0009-0008-1291-2281
AD  - Senior Lecturer, Faculty of Architecture, Building and Planning, The University of Melbourne, Melbourne, Australia, shang.gao@unimelb.edu.au, orcid.org/0000-0002-4161-5592
AD  - Senior Lecturer, Faculty of Architecture, Building and Planning, The University of Melbourne, Melbourne, Australia, mehran.oraee@unimelb.edu.au, orcid.org/0000-0002-5928-4760
ED  - Hamzeh, Farook
ED  - Poshdar, Mani
ED  - Garcia-Lopez,, Nelly P.
AB  - The construction industry continues to face significant performance challenges due to  defective work, with rework estimated to account for a substantial portion of project costs. Despite the adoption of Lean Construction principles, effective learning from defects remains hindered by fragmented workflows and the unstructured nature of defect data. Furthermore, while Artificial Intelligence (AI) has shown promise in quality management, existing research predominantly focuses on Computer Vision for reactive detection rather than leveraging textual data for proactive prevention. To address this gap, this study adopts a conceptual research methodology to propose the Lean AI-Driven Defect Management Framework. This framework integrates the philosophical pillars of Lean Construction: Respect for People, Flow, and Value, with the analytical capabilities of Large Language Models (LLMs). The proposed model conceptualises defect data not as a compliance metric, but as a strategic resource for organisational learning. By utilizing LLMs to process unstructured defect records and identify recurring patterns, the framework supports "just-in-time" decision-making and targeted Lean interventions, such as strategic "Muda walks." This study contributes to the field by offering a theoretical foundation for integrating human-centric Lean values with AI-driven knowledge systems, shifting defect management from reactive correction to continuous, data-driven improvement.
KW  - Lean construction
KW  - defect management
KW  - artificial intelligence
KW  - large language model
KW  - Lean Construction 4.0.
PB  - 
T2  - Proceedings of the 34th Annual Conference of the International Group for Lean Construction (IGLC 34)
DA  - 2026/06/22
CY  - Singapore, Singapore
L1  - http://iglc.net/Papers/Details/2507/pdf
L2  - http://iglc.net/Papers/Details/2507
N1  - Export Date: 19 June 2026
DB  - IGLC.net
DP  - IGLC
LA  - English
ER  -