Bridge Condition Assessment Based on Image Data and Digital Twins

Many different approaches using modern digital technologies were recently developed to support engineers with the acquisition of visual inspection data, such as the usage of small unmanned aircraft systems (UAS) equipped with high-quality cameras. The images obtained are used, amongst others, for photogrammetric reconstruction methods or image-based anomaly detection, which leads to a high potential of automation in condition assessment, reducing time and costs. This article presents approaches for the integration of image-based inspection data sets into an automated workflow towards condition rating of damaged infrastructures. To this end, it is shown how 3D annotations are combined with information from a digital twin, such that further properties are assigned to the detected structural anomalies, in order to enrich the digital twin. Finally, the proposed methods are applied to a case study to show the feasibility in a practical use case.