CASE STUDY FOR UAS-ASSISTED BRIDGE INSPECTIONS

Abstract. Bridge inspections are typically expensive and time-consuming, especially in regards of the inspection difficult-to-reach areas. In recent years, unmanned aerial systems (UASs) have gained attention due to their flexible data acquisition. However, UAS generate large quantities image video data, which currently analysed manually. Additionally, identified damages not assessed accurately geometric characteristics location. this paper, we propose a time-effective framework for UAS-based bridge methodology that combines 3D information from photogrammetry machine learning based object detection allow direct measurements images. Concretely, use two-step flight planning reconstruct using limited manual effort. Second, detect frequently occurring such as exposed rebars concrete spalling on imagery. Finally, spatial location imagery significantly improve results geolocate them. We evaluate our proposed decommissioned bridge. The trained YOLOv8 models prove capable transfer both own online benchmarks. photogrammetric reconstruction also proves be sufficiently reliable. Overall, these first steps automating routine provide crucial evidence continue developing method.