Advanced monitoring of road pavement and infrastructure degradation using mobile laser scanning and ground penetrating radar

The primary objective of this study is to develop efficient surveying and data processing methodologies that enable the extraction of more detailed metric data on road infrastructure than what can be obtained through traditional survey techniques.

Condition assessments conducted using traditional methodologies may be risky and, in some cases, ineffective. The Mobile Laser Scanner (MLS) technique, based on LiDAR technology, is widely adopted as a reliable alternative, as it allows for the generation of dense, accurate point clouds of both the road surface and associated artworks.

The aim of our work is to provide a comprehensive workflow for processing MLS data to generate useful indicators that describe the functional and structural characteristics of both the pavement and related structures, with the goal of optimizing decision-making processes for infrastructure managers.

Data processing for road surface analysis involves three main stages: (i) Extraction of points corresponding to the road pavement or the surfaces of associated structures; (ii) Generation of a curvilinear abscissa Digital Elevation Model (DEMc); and (iii) Analysis of surface regularity and the intrados of the artworks.

Point cloud filtering relies on the M-estimator SAmple Consensus (MSAC) algorithm, a robust variant of the RANSAC method. The DEMc is designed to follow the curvilinear alignment of the road axis. A curvilinear planimetric grid is first generated, with the curvilinear abscissa corresponding to the points marked by horizontal road signage. Elevation values are then assigned to each grid node, derived via local interpolation of points from the road surface. Surface condition assessment and cross-slope analysis are conducted by examining each cross-profile extracted from the DEMc. For each profile, regularity indices such as Rut Depth, as well as characteristic geometric parameters like transverse slopes, are calculated.

Regarding the analysis of structures, particularly tunnel intrados, our study proposes a methodology that utilizes an automatic unrolling algorithm for point clouds of the intrados, based on the RANSAC method. Intensity values of the LiDAR data are then analyzed to detect potential water infiltrations, while roughness values are calculated to assess surface integrity and identify cracks or steel bar ejections. The results, though focused on only two types of degradation, are useful for pinpointing tunnel sections in need of urgent intervention, thereby indicating areas of high priority for action or alert. The entire process is implemented in MATLAB.

The condition of the underlying layers of the road pavement were examined through Ground-Penetrating Radar (GPR) measurements to identify potential damage sources responsible for the deterioration of the surface layers. This also enables an assessment of whether the deformations affect only the superficial layers or extend to deeper strata. The outcome of the entire process is the creation of an Atlas in QGIS.

Data acquisition was carried out using a Leica Pegasus TRK500 Neo MLS, in collaboration with C.U.G.RI., Leica Geosystems for the survey, and SPN Salerno Pompei Napoli S.p.A. for logistical support. The survey was conducted over a 4 km stretch of the A3 highway (Campania Region, Italy), an area significantly affected by hydrogeological hazard.