Modelling Techniques for Occupational Health and Safety Management in Construction: A Systematic Literature Review

In the past several decades, the construction industry has been increasingly aware of the importance of occupational health and safety (OHS) management, owing to its currently appalling OHS performance (i.e. high fatalities and injuries rate). To improve OHS performance, careful and holistic safety planning is required throughout the entire construction life cycle, which would typically span across design-to-construction, operation-and-maintenance, decommissioning and/or renewal. Specifically, safety planning includes the processes of identifying potential hazards in construction, analysing the corresponding risks, and selecting appropriate safety measures to mitigate or eliminate such risks. Recent advancements in information technology have enabled the application of various modelling techniques to simplify and optimise the safety planning processes. However, there is limited research on analysing how different modelling techniques have been used in tackling different OHS issues at different stages of the construction life cycle. This knowledge gap could prevent researchers from selecting appropriate modelling techniques for improving safety planning. This study aims to bridge the gap by conducting a literature review in the construction OSH literature published in the last five years. The study found four classes of modelling techniques are dominant: 1) Structural Equation Modelling (SEM); 2) Agent-Based Modelling (ABM); 3) Building Information Modelling (BIM); and 4) Machine Learning (ML) or Deep Learning (DL). It also analysed the extent to which the four identified modelling techniques were used for planning OHS within different stages of the construction life cycle. Additionally, the limitations of each technique were examined.