Identifying falling-from-height hazards in building information models: From the perspectives of time and location

Falling-from-Height (FFH) is the leading hazard among the 'Fatal Five' in the architectural/engineering/construction (AEC) industry, often resulting in severe consequences. Therefore, preventing these accidents is critical at the earliest stage. However, current FFH hazard identification methods suffer from a high workload, computational intensity, and lack of comprehensiveness. We propose a voxelization-based method to identify the temporal and spatial information on FFH hazards in Building Information Modeling (BIM). This method considers the falling, accessibility, and temporal conditions. We compared this method with a traditional FFH hazard identification method. The results indicate that the proposed method can rapidly identify all potential FFH hazard information, with a recall of 100 % and a precision of 81.5 %. It is 30 times faster than the offset geometry method and is not affected by the curvature of model surfaces, improving the effectiveness of FFH hazard identification. The proposed method is highly suitable for preventing FFH accidents and can be implemented during the design and construction preparation stages, increasing the effectiveness of FFH accident control. • The proposed method can quickly and efficiently identify all potential FFH hazards by considering temporal and spatial information. • The proposed method considers the elements’ protections against FFH hazards, avoiding false positives (FPs). • The proposed method has the ability to visualize the output facilitates preventing FFH hazards.