PSI - Issue 64

Konrad Bergmeister et al. / Procedia Structural Integrity 64 (2024) 14–20 Konrad Bergmeister / Structural Integrity Procedia 00 (2019) 000 – 000

18

5

Fig. 2. (left) UAV-obtained image of the base of a bridge pier; (right) AI-enhanced image with two automatically added annotations referring to cracks (red) and spalling (blue). Courtesy of: Archetipo srl & A. Flora (amedeo.flora@unibas.it), D. Cardone (donatello.cardone@unibas.it); School of Engineering, University of Basilicata, Potenza, Italy 3. From point clouds to BIM models and risk assessments 3.1. Developing BIM models from point clouds The use of UAV effectively delivers a large quantity of photogrammetric data regarding the external surfaces of a bridge together with a geographic information system (GIS) catalogue. Collecting such databases, commonly referred to as point clouds, is the basis for automatically deriving 3D BIM models (Pallante et al. 2024). The goal of utilising a BIM bridge model in the current work is to deliver a digital repository that visualises the geometry of the bridge's structural components to be inspected and effectively demonstrates the AI-detectable defects of the structure. For this reason, technical drawings are combined with the point cloud data, resulting in accurate 3D digital models of the main components of the bridge, including piers, girders, abutments, bearings, etc. Following the example of the bridge pier in Fig. 2, the proposed methodology allows for the automatic representation of the specific component in the 3D digital model shown in Fig. 3 (left) and the BIM model shown in Fig. 3 (right).

Fig. 3. (left) The 3D digital model of the pier with the defects; (right) The BIM model of the pier with the defect. Courtesy of: Archetipo srl & A. Flora (amedeo.flora@unibas.it), D. Cardone (donatello.cardone@unibas.it); School of Engineering, University of Basilicata, Potenza, Italy