A Dynamic Fit-Out Scheduling Framework for Digital Twin-Enabled Modular Integrated Construction

Modular Integrated Construction (MiC) is a novel construction method for its approved space-saving and time-saving for the construction industry. Fit-out construction, as the last essential process of MiC module assembly, significantly affects the delivery of MiC modules for its complexity and flexibility in operations. Emerging technologies, such as the Internet-of-Things (IoT) and Digital Twin (DT), capture real-time fit-out operation data and worker information, which reshape the fit-out job scheduling problem. This study proposes a dynamic fit-out scheduling framework by integrating job scheduling and worker assignment for DT-enabled MiC. First, this study defines the workflow of fit-out operations and the fit-out scheduling problem in MiC. The new fit-out scheduling problem considers job scheduling and worker assignment simultaneously. Second, this work designs a DT-enabled framework for cyber-physical operational synchronisation in MiC regarding workers, operations and jobs. Third, this work implements a dynamic fit-out scheduling method based on Deep Reinforcement Learning. This work intends to conduct experimental analysis using real-life case data. This work aims to provide a real-time data-driven solution for fit-out scheduling problems in MiC.