Construction projects often suffer from delays, cost overruns, and fragmented control due to isolated implementation of Computer Vision (CV), Internet of Things (IoT), Building Information Modelling (BIM), and Machine Learning (ML). This study addresses this gap by proposing a comprehensive Digital Twin framework that integrates these technologies into a unified system for real-time monitoring and predictive control of labor, material, equipment, and activity (LMPA).
A four-phase bibliometric approach was followed: (1) retrieving 2,032 studies (2014–2024) from the search database; (2) screening through multi-level filtering to retain 534 relevant papers; (3) network analysis and mapping keywords in Gephi and forming thematic clusters using the Louvain algorithm; and (4) identifying the research gap of fragmented CV, IoT, BIM, and ML applications and developing a Digital Twin framework for real-time LMPA control.
Nineteen functional clusters were identified across the four domains: CV (Visual Understanding, Edge Hardware, Metrics, LMPA Algorithms, Analytics), IoT (Sensors, Transmission, Edge Devices, Signal Processing, Dashboards), BIM (Foundations, 4D/5D Control, Progress Tracking, As-Built Alignment), and ML (Data Preparation, Model Training, Forecasting, Optimization, Real-Time Feedback). These form a Digital Twin framework for real-time, closed-loop project monitoring and control.
Unlike prior reviews focusing on single technologies or static visualisation, this study integrates CV, IoT, BIM, and ML into a single, evidence-based Digital Twin framework tailored to closed-loop LMPA control.
