PSI - Issue 62

Antonio Di Pietro et al. / Procedia Structural Integrity 62 (2024) 755–762 Antonio Di Pietro et al. / Structural Integrity Procedia 00 (2019) 000 – 000

756

2

1. Introduction Great efforts have been recently made to address issues related to infrastructure safety, which have escalated due to the collapse of some bridges (Clemente, 2020). It is cost-effective to intervene promptly in case of significant defects, damage, or deterioration to avoid economic burden and prevent further damage that could affect the bridge lifespan. Bridges are fundamental parts of any transport infrastructure, linked to public safety and economic development, requiring ongoing maintenance and public spending. In the recent Italian guidelines (LG2020, 2020), the criteria for the safety assessment of bridges have been defined (Ormando et al. 2022). These are initially based on census, i.e., the collection of relevant information about the structure, and visual inspection to identify defects and deterioration. Following the inspection, bridges are rated according to a Class of Attention (Buffarini et al. 2022, 2023). Nowadays, accurate and sophisticated procedures for static and dynamic monitoring are available and commonly used (Bongiovanni et al. 2021). However, visual inspection is still a fundamental component for the evaluation of the structural health of a bridge. The purpose is to identify any damage or deterioration, such as cracks, corrosion, deformations, scour, or damage caused by impacts. Binoculars and high-resolution cameras are often used to examine hard-to-reach parts of the bridge. Photographic documentation, as well as defect rating, are essential to assess the health status of the bridge. However, visual inspections present logistical and technical challenges. Procedures are burdensome in terms of time and resources and can be risky for inspectors, especially in emergency situations. They often require aerial platforms, such as by-bridges or scaffolding, as well as specialized personnel. Several bridge components, such as elevated parts of the superstructure or piers, are typically difficult to access. Consequently, the use of Unmanned Aerial Systems (UAS), also called drones, is on the rise, showing significant efficacy, especially in emergency scenarios. The deployment of drones is proving to be complementary to traditional methods and, in the near future, is likely to become a valid alternative to them (Kim et al. 2022, Mandirola et al. 2022). UASs allow to operate in confined spaces, perform close-range flights, and acquire high-definition data from challenging angles. By using specialized software, the data collected from drones can be transformed into digital 3D models, valuable for inspections and structural assessments. Therefore, photogrammetry emerges as a cost-effective option compared to remote sensors like laser scanners, albeit with some limitations (Gaspari et al., 2022). The conversion of this data into Finite Element or BIM models represents a rapidly evolving sector. In this paper, a practical approach to the inspection and modeling of bridges using UAS is proposed. Various aspects are examined, from aerial photogrammetry to the use of multirotor UAS for data acquisition. The Beroide Bridge, a 100-m long five-span bridge with a variable height of about 5 m from the bed of the Maroggia river at Spoleto, Italy, is taken as case study. A georeferenced photogrammetric survey was performed, generating a dense point cloud, followed by the creation of a digital twin. 2. UAS Aerial Photogrammetry Workflow In order to have an initial knowledge of the geographic context in which the bridge is located, a visual inspection of the area can be conducted via Google Earth. This also allows to get information on access to the structure, drone to use and type of flight to execute. A preliminary check should also be made on the Air Navigation Service Provider (ANSP) to check if flights are allowed according to country-specific regulations. The bridge accessibility should be examined through an on-site preliminary inspection, which helps identifying any obstacles not present in an online inspection setting. Aerial photogrammetry using drones entails a series of precise steps that must be meticulously followed to achieve accurate results (Figure 1). The main steps are described in the following. 2.1. Flight Planning The area to be mapped must first be defined establishing the flight plan, including altitude, trajectory, and intervals for photo capture. A critical feature is the ability to execute pre-programmed flights, designed with dedicated photogrammetric capture software. The proposed method for data acquisition is to segment the aerial survey into two distinct parts. The initial segment involves a zenithal flight, capturing comprehensive images of the entire bridge and its surroundings. This flight plan can be set up using specialized software and allowing the pre-setting of key photogrammetric parameters, such as image overlaps and Ground Sampling Distance (GSD), i.e., the ground