Artificial intelligence (AI) and machine learning in building information modeling (BIM)-based construction cost estimation: a systematic review

Construction cost estimation is a multifaceted process crucial for effective project planning and management. Leveraging building information modeling (BIM) as a pivotal source of information, coupled with the integration of artificial intelligence (AI) and machine learning (ML) techniques, presents a transformative potential for automating the intricate stages of cost estimation in the construction industry. Accordingly, this paper presents a systematic literature review, scrutinizing 77 papers encompassing different facets of construction cost estimation while integrating AI techniques into one or more stages, with BIM as a foundational dataset. The review begins by addressing the building breakdown and classification of components, a critical step in the semantic enrichment of BIM 3D digital models. It then examines subsequent processes, including the preparation of quantity take-off (QTO) by quantifying 3D components and identifying corresponding materials. Finally, it considers comprehensive cost estimation, which involves either calculating direct total project costs or aggregating costs for materials, labor, and equipment. The paper evaluates the current landscape of AI-BIM integration, identifies prevailing challenges and limitations, and propounds insightful future directions. By delving into the complexities of AI-driven cost estimation within the 3D context of BIM, this research contributes significantly to the ongoing evolution of data-driven practices and 3D digital modeling in construction management and urban studies.