Proof-of-Concept Study for Model-Based Construction Safety Diagnosis and Management Driven by Prevention through Design

The primary objective of this research is to demonstrate the feasibility of a model-based construction safety assessment system using building information modeling (BIM) and diagnosing accident-prone BIM objects through prevention through design (PtD). Although extensive research has focused on early risk detection and accident predictions in safety, previous approaches have often missed opportunities to identify safety issues arising from design choices. Potential safety risks have been assessed retrospectively by reconstructing safety concerns based on completed design options. To address this gap, this research aims to provide foresight regarding construction safety hazards from the early design stage. First, risks embedded in design decisions are identified by analyzing safety incident reports using text-mining techniques. Then, the relationships among design elements, accident precursors, and risk events are established through if-then relationships. The potential hazards associated with design choices are evaluated by developing and running visual scripts and assessing design model parameters in BIM. This approach enables architects to track construction risks during their design stage, even without extensive onsite construction experience. In addition, owners can evaluate design decisions considering construction safety risks, and contractors can forecast and monitor risky elements, materials, and locations during construction execution. The research outcomes contribute to enhancing safety risk awareness in the early design phases and support efficient and predictive safety management during construction.