PSI - Issue 84

Antonio Di Pietro et al. / Procedia Structural Integrity 84 (2026) 57–64

59

integration of multi-return airborne LiDAR with a multi-view flight strategy, (ii) the calibration and strip-adjustment workflow tailored to tall-pier geometries, and (iii) the derivation of structurally meaningful as-built models suitable for BIM/FEM and long-term monitoring.

Fig. 1. Workflow of the LiDAR-based UAS methodology.

2. UAS-LiDAR Survey Methodology 2.1. Preliminary Analysis and Flight Planning

A preliminary reconnaissance of the bridge and surrounding terrain is conducted to analyse the structural layout, local topography, and environmental constraints. This includes: 1) identification of primary and alternate take-off and landing sites; 2) verification of regulated airspace, operational limitations, and safety perimeters; 3) recognition of obstacles such as vegetation, pylons, and elevated elements; and 4) definition of logistical support areas for mission control and equipment staging. When needed, an initial photogrammetric survey may be acquired to update the local Digital Surface Model (DSM) and enhance terrain-following flight planning. This enables accurate definition of flight altitudes, scan geometry, and obstacle-avoidance parameters. LiDAR-based acquisitions generally require flexible flight altitudes to ensure correct viewing angles of structural elements such as deck soffit, pier heads, or abutments. If over-road flights cannot be authorised, alternative trajectories, potentially conducted during low-traffic periods or after a site-specific risk assessment, can be adopted. Since LiDAR sensors are insensitive to illumination, night-time operations may represent a viable option when daytime constraints prevent direct line-of-sight access to structural components or when traffic cannot be interrupted. The configuration of the LiDAR payload is defined according to the expected operating ranges, the required point density, and the geometrical complexity of the bridge environment. Key acquisition parameters, such as pulse repetition rate, scan frequency, scan angle, and number of returns, are selected in a coordinated manner, since they jointly determine the density, completeness, and angular distribution of the point cloud. In environments characterised by tall vegetation or partial occlusions, the use of multiple returns significantly improves the probability of capturing canopy-filtered terrain points along with structural elements located beneath shaded or obstructed regions. Before data acquisition, an accurate IMU–LiDAR boresight calibration is performed to ensure proper alignment between the inertial and laser reference frames. This step is essential because small misalignments may propagate along the flight trajectory and generate systematic biases in the final point cloud, particularly when reconstructing slender components such as pier shafts, diaphragms, or deck soffit. Whenever available, the UAS platform is connected to a Global Navigation Satellite System (GNSS) - i.e., satellite-based positioning systems such as GPS, Galileo, or GLONASS—using a Real-Time Kinematic (RTK) correction service to obtain centimetric positioning accuracy during flight. However, due to typical constraints of bridge environments, such as deck obstruction, steep embankments, narrow canyons, or dense vegetation, short-term loss of correction data may occur. For this reason, the processing workflow includes the option of performing a Post-Processing Kinematic (PPK) refinement, which 2.2. Sensor Configuration and Calibration