Automatic Fall Risk Identification using Point Cloud Data in Construction Excavation

Safety continues to be among the top issues in the construction industry after experiencing 775 fatalities in 2012. Among all those fatalities, fall has been considered as one of the leading contributors for several years. This paper presents a method that automatically identifies fall risks on construction sites under excavation by utilizing laser scanning technology. It first extracts safety rules that are related to fall risks from OHSA standards and construction best practices. Then, it collects sets of point cloud data of a construction site under excavation using laser scanning, registering and cleaning the point cloud data afterwards. Finally, it develops an identification algorithm based on those rules and applies the algorithm onto the data to identify potential fall risks by analyzing geometrical properties. An experimental trial is also conducted in this paper and results show that the method successfully identifies those fall risks. The presented method can actively monitor the fast changing situations of construction sites under excavation and provide inspectors and project managers with valuable information about fall risks, helping them make good safety decisions and prevent fall accidents and fatalities.