Modeling job sites in real time to improve safety during equipment operation

Abstract

Real-time three-dimensional (3D) modeling of work zones has received an increasing interest to perform equipment operation faster, safer and more precisely. In addition, hazardous job site environment like they exist on construction sites ask for new devices which can rapidly and actively model static and dynamic objects. Flash LADAR (Laser Detection and Ranging) cameras are one of the recent technology developments which allow rapid spatial data acquisition of scenes. Algorithms that can process and interpret the output of such enabling technologies into threedimensional models have the potential to significantly improve work processes. One particular important application is modeling the location and path of objects in the trajectory of heavy construction equipment navigation. Detecting and mapping people, materials and equipment into a three-dimensional computer model allows analyzing the location, path, and can limit or restrict access to hazardous areas. This paper presents experiments and results of a real-time three-dimensional modeling technique to detect static and moving objects within the field of view of a high-frame update rate laser range scanning device. Applications related to heavy equipment operations on transportation and construction job sites are specified.