Publications / 2024 Proceedings of the 41st ISARC, Lille, France
This study merges artificial intelligence (AI) image recognition technology with Building Information Modeling (BIM), to develop a prototype system for the automation and visualization of construction site progress control. Our focus is the strategic deployment of multiple construction site surveillance cameras using a BIM model to encompass the entirety of the construction site. Following the capture of camera images, the application of object detection techniques within AI image recognition locates all actively constructed objects in the images, subsequently identifying the construction phases to which these objects belong. By integrating multiple camera perspectives from the site into a BIM model, the results of AI detection are automatically inputted into the corresponding components of the model. Finally, real-time on-site progress information obtained from the BIM model is compared with the progress schedule, and the comparative results are visually presented on the BIM model components in distinct colors. Through this visual approach, managerial personnel can intuitively and instantly control the construction progress.