AUTOMATED RECOGNITION OF WORK PROGRESS AT A CONSTRUCTION SITE

The progress of construction process could be obtained through the records by construction managers and subcontractors' daily records. However, as those records are based on human observation, there is no objective data to support their credibility, and human errors might exist on them. Currently, it has become relatively easy to grasp the situation at a construction site as the point clouds obtained by 3D laser scanners as well as by the photogrammetry method. The authors have developed methods for identifying components placed at the site by comparing point clouds with 3D CAD model. This paper presents the methods, and verifies these validity through the experiments to identify precast concrete components automatically based on point clouds.