Automatic mapping of schedule activities and reality models for tracking construction progress

Purpose – The purpose of this research is to develop a novel method for automatically aligning schedule activities with reality models to track construction project progress. This approach aims to overcome the limitations of traditional methods that depend on 4D Building Information Models (BIM), which are often labor-intensive to create and can quickly become outdated. Design/methodology/approach – The proposed methodology utilizes visual data to monitor construction progress. It employs 3D BIM or ground control points to align reality models and utilizes point cloud segmentation and image segmentation for progress detection. Natural language processing (NLP) is used to extract location, object and task information from schedule activities. A distance-based matching technique is applied to map reality model components with the corresponding scheduled activities. The method is tested on two building construction projects: one with a 3D BIM and another without. Findings – The testing on two case studies demonstrated the method's capability to automatically track construction progress by accurately aligning schedule activities with reality models, even in the absence of an up-to-date 4D BIM. The approach showed significant potential in streamlining progress monitoring processes, enhancing accuracy and reducing the need for manual updates, thereby supporting more effective project management and decision-making. Originality/value – This research presents a novel integration of NLP, point cloud segmentation and image segmentation for construction progress monitoring, offering a unique solution that bypasses the need for a current 4D BIM. The approach addresses key industry challenges by automating the alignment of schedule activities with reality models, providing a valuable tool for enhancing the efficiency and accuracy of project progress tracking in construction management.