Taiwo, R, Bello, I.T, Abdulai S.F, Yussif, A, Salami, B.A, Saka, A.B., Ben Seghier, M and Zayed, T 2025. Generative artificial intelligence in construction: A Delphi approach, framework, and case study. Alexandria Engineering Journal. 116, pp. 672-698. https://doi.org/10.1016/j.aej.2024.12.079
| Title | Generative artificial intelligence in construction: A Delphi approach, framework, and case study |
|---|---|
| Type | Journal article |
| Authors | Taiwo, R, Bello, I.T, Abdulai S.F, Yussif, A, Salami, B.A, Saka, A.B., Ben Seghier, M and Zayed, T |
| Abstract | The construction industry plays a crucial role in the global economy, contributing approximately $10 trillion and employing over 220 million workers worldwide, but encounters numerous productivity challenges with only 1 % annual growth compared to 2.8 % for the global economy. These challenges span various processes, including design, planning, procurement, inspection, and maintenance. Generative artificial intelligence (GenAI), capable of producing new and realistic data or content such as text, images, videos, or code from given inputs or existing knowledge, presents innovative solutions to these challenges. While there is an increasing interest in the applications of GenAI in construction, a detailed analysis of its practical uses, advantages, and areas ripe for development is still evolving. This study contributes to this emerging area by offering an insightful analysis of the current state of generative AI in construction. It has three objectives: (1) to identify and categorize the existing and emerging generative AI opportunities and challenges in the construction industry via a Delphi study; (2) to propose a framework enabling construction firms to build customized GenAI solutions; and (3) to illustrate this framework through a case study that employs GenAI model for querying contract documents. Through systematic review and expert consultation, the study identified 76 potential GenAI applications across construction phases and 18 key challenges distributed across domain-specific, technological, adoption, and ethical categories. The case study's findings show that retrieval augmented generation (RAG) improves the baseline large language model (LLM), GPT-4, by 5.2, 9.4, and 4.8 % in terms of quality, relevance, and reproducibility. The study recommends a structured approach to GenAI implementation, emphasizing the need for domain-specific customization, robust validation protocols, and careful consideration of ethical implications. This study equips academics and construction professionals with a comprehensive analysis and practical framework, facilitating the integration of GenAI techniques to enhance productivity, quality, safety, and sustainability across the construction industry. |
| Journal | Alexandria Engineering Journal |
| Journal citation | 116, pp. 672-698 |
| ISSN | 2090-2670 |
| 1110-0168 | |
| Year | 2025 |
| Publisher | Elsevier |
| Publisher's version | License CC BY 4.0 File Access Level Open (open metadata and files) |
| Digital Object Identifier (DOI) | https://doi.org/10.1016/j.aej.2024.12.079 |
| Publication dates | |
| Published | 09 Jan 2025 |