Computing 3D blind spots of construction equipment: Implementation and evaluation of an automated measurement and visualization method utilizing range point cloud data

Abstract

Limited visibility due to blind spots of construction equipment was responsible for 55% of the visibility-related fatalities in the construction industry. Knowledge of blind spot aids in improving safety on construction sites and the design of equipment cabin itself. Existing approaches for blind spot measurement that follow international standards typically require a time-consuming set-up and are limited by the number of different visibility analyses that can be performed. A new approach to compute blind spot in a fast and efficient way using point cloud data of equipment is presented. The developed approach allows performing different analyses such as: volumetric blind spots, blind spot map, 12 m circumference visibility, rectangular 1 m boundary visibility, and worker visibility. In addition, the above set of analyses can be performed from different viewpoints located virtually anywhere inside the equipment cabin. Validating against synthetic “noisy” point clouds, robustness and accuracy of the approach is established.