Case study on accuracy comparison of digital twins developed for a heritage bridge via UAV photogrammetry and terrestrial laser scanning

Over the last decade, advanced remote technologies have been considerably utilised for digitisation of civil infrastructure assets, particularly bridges, and providing accurate data for indirect inspection and assessment through digital twins of their physical counterparts. Although advanced emerging technologies such as Unmanned Aerial Vehicles (UAVs) photogrammetry and Terrestrial Laser Scanning (TLS) established a suitable alternative against labour-intensive and expensive methods of direct inspections, the research is still lacking a comparative analysis of accuracy and reliability of the associated digital twins. This paper serves to investigate and evaluate the geometric accuracy of two 3D reality models generated from an existing heritage bridge in Australia extracted from both UAV-based photogrammetry and TLS based point clouds. The comparative results show the capability of both UAV photogrammetry and TLS based point clouds for bridge inspection; however, TLS expresses a higher level of points’ density with suitable accuracy subject to implementation of precise as-built 3D reconstruction method