Towards automated construction progress monitoring using BIM-based point cloud processing

On-site progress monitoring is essential for keeping track of the ongoing work on construction sites. Currently, this task is a manual, time-consuming activity. The research presented here, describes a concept for an automated comparison of the actual state of construction with the planned state for the early detection of deviations in the construction process. The actual state of the construction site is detected by photogrammetric surveys. From these recordings, dense point clouds are generated by the fusion of disparity maps created with Semi-Global-Matching (SGM). These are matched against the target state provided by a 4D Building Information Model (BIM). For matching the point cloud and the BIM, the distances between individual points of the cloud and a component's surface are aggregated using a regular cell grid. For each cell, the degree of coverage is determined. Based on this, a confidence value is computed which serves as basis as for the existence decision concerning the respective component. Additionally, process-and dependency-relations are included. First experimental results from a real-world cases study are presented and discussed.