Real-time Spatial Detection and Tracking of Resources in a Construction Environment

Construction accidents with heavy equipment and bad decision making can be based on poor knowledge of the site environment and in both cases may lead to work interruptions and costly delays. Supporting the construction environment with real-time generated threedimensional (3D) models can help preventing accidents as well as support management by modeling infrastructure assets in 3D. Such models can be integrated in the path planning of construction equipment operations for obstacle avoidance or in a 4D model that simulates construction processes. Detecting and guiding resources, such as personnel, machines and materials in and to the right place on time requires methods and technologies supplying information in real-time. This paper presents research in real-time 3D laser scanning and modeling using high range frame update rate scanning technology. Existing and emerging sensors and techniques in three-dimensional modeling are explained. The presented research successfully developed computational models and algorithms for the real-time detection, tracking, and threedimensional modeling of static and dynamic construction resources, such as workforce, machines, equipment, and materials based on a 3D video range camera. In particular, the proposed algorithm for rapidly modeling three-dimensional scenes is explained. Laboratory and outdoor field experiments that were conducted to validate the algorithm’s performance and results are discussed.