Digital Twin Enabled Construction Site Monitoring (CSM) Method with Edge-Cloud Collaboration

In order to realize efficient construction site safety management, this paper proposes a digital twin enabled construction site monitoring (CSM) method with edge-cloud collaboration. Firstly, the high-fidelity virtual model including the building information model, physical model, behavior model and rule model is constructed to realize the accurate mapping of construction sites from physical space to virtual space and the causes analysis between behaviors and construction accidents. Secondly, the edge-cloud collaborative architecture is also designed to realize the reasonable allocation of computing tasks. In the edge-layer, the industrial CCD cameras and deep learning-based construction site detection methods are applied to realize the real-time images collection and detection on the construction sites. In the cloud layer, the cloud computing technology is used to realize the efficient data management, analysis, and model training. Finally, the experiment analysis is also implemented for the performance of the proposed deep learning-based construction site monitoring methods.