Integrating knowledge management and large language models to advance construction Job Hazard Analysis: A systematic review and conceptual framework

Conducting a Job Hazard Analysis (JHA) remains essential for managing safety risks in construction; however, the process is often manual, subjective, and knowledge-intensive. While numerous studies have proposed tools and techniques to enhance JHA, a comprehensive synthesis through the lens of construction safety knowledge management (CSKM) has been lacking. This systematic review fills that gap by: (1) Critically examining recent advancements in JHA practices with a focus on how tacit and explicit safety knowledge is acquired, integrated, and applied; (2) Analyzing the emerging role of interoperable and semantic technologies – such as building information modeling (BIM), ontologies, knowledge graphs (KGs), and semantic reasoning – in supporting JHA through CSKM; and (3) Proposing a novel conceptual framework that outlines the potential integration of large language models (LLMs) to automate and enhance JHA processes. Using the preferred reporting items for systematic reviews and meta-analyses (PRISMA) methodology, 90 peer-reviewed studies were systematically reviewed and thematically analyzed. The results reveal actionable patterns in how digital technologies and knowledge management strategies are converging to address longstanding issues in hazard identification and decision-making. By embedding institutional knowledge into LLM-supported CSKM, this review contributes to developing safer, more adaptive, and ultimately more sustainable construction practices.