VR and Deep Learning Based Digital Twin System for Visual Inspection Training of Road Bridge

Road bridges constitute critical infrastructure that ensures safe and efficient transportation. However, they are susceptible to damage caused by aging and external factors. For example, cracking, corrosion of members, surface delamination and various other types of deterioration progress depending on the material and structural type. Beyond the degradation of structural functionality and durability, such damage heightens risks to vehicles and pedestrians. Consequently, effective inspection and maintenance strategies are imperative. In this paper, we propose a digital twin system based on virtual reality and deep learning to train inspection engineers for various bridge types and maintenance management. The proposed system enables bridge inspection training within a virtual environment using a three-dimensional bridge model and provides real-time feedback to the trainees. Furthermore, the effectiveness of the proposed system is evaluated through its application to a road bridge on National Route 2, a critical transportation artery in Japan with high traffic volumes.