Automatic Mountain Tunnel Invert Excavation with Digital Twin Management in Non-GNSS Environments

Abstract: In non-GNSS environments, such as mountain tunnel construction sites, high-precision position estimation is challenging. This research proposes a real-time tracking method using LiDAR to capture reference points as a positioning technique for non-GNSS environments. Using this to acquire backhoe coordinates during invert excavation enabled the automation of rough excavation for the invert. Furthermore, by providing an overview visualization of the processes before and after automated construction, the surrounding conditions, and related tasks on a digital twin, it demonstrated safe automated construction management from remote locations. This research contributes to the automation and unmanned operation of invert excavation in mountain tunnels.