Human–Machine Collaborative Management of Pre-Embedded Components for Submerged Tunnel Segments Based on BIM and AR

In submerged tunnel construction, the installation accuracy of pre-embedded components directly impacts subsequent engineering quality and operational safety. However, current on-site construction still primarily relies on manual measurement and two-dimensional drawings for guidance, resulting in significant positioning errors, delayed information transmission, and inefficient installation inspections. To enhance the digitalization and intelligence of submerged tunnel construction, this paper proposes a BIM- and AR-based human–machine collaborative management method for pre-embedded components in submerged tunnel segments. This method establishes a site-wide panoramic model as its foundation, enabling intelligent matching of component model geometry and semantic information. It facilitates human–machine interaction applications such as AR-based visualization for positioning and verification of pre-embedded components, information querying, and progress simulation. Additionally, the system supports collaborative operations across multiple terminal devices, ensuring information consistency and task synchronization among diverse roles. Its application in the Mingzhu Bay Submerged Tunnel Project in Nansha, Guangzhou, validates the feasibility and practical utility of the proposed workflow in a pilot case, and indicates potential for further validation in broader construction settings.