PSI - Issue 84

Alessandro Scala et al. / Procedia Structural Integrity 84 (2026) 489–496

490

1. Introduction Landslide risk, one of the most frequent causes of failure (Scala et al., 2025; Gabrieli et al., 2025), represents one of the most delicate aspects in the safety assessment of bridges and viaducts. These assessments, together with the design of the mitigation measures required to safeguard the Italian infrastructure network, became mandatory and standardized following the collapse of the Morandi Bridge in 2018. The evaluation covers four types of risk: structural, seismic, hydraulic and, specifically, landslide risk. While the procedures defined by the national Guidelines (MIMS, 2022) are clear and well-established for the first two, supported by tools such as the so-called “defect sheets” (Niero et al., 2025), the same level of clarity is not found when dealing with hydro-geological hazards. The assessment of landslide risk leaves considerable room for interpretation by the appointed engineer. Since these phenomena depend strongly on rainfall intensity, on subsurface stratigraphy, which is often difficult to characterize, and on geomorphological conditions that may vary significantly from site to site, accurately estimating potential paths, volumes, and movement velocities is challenging and time-consuming. This becomes particularly critical in the initial stages of applying the Guidelines: the inventory phase, aimed at collecting all available information regarding the structure and its surroundings; the inspection phase, whose purpose is to confirm or refine the previously gathered data; and the subsequent assignment of the risk attention class for the structure (Levels 0, 1, and 2). These levels are primarily intended to identify the structures that exhibit the highest risk and therefore require more detailed analyses in the subsequent phases, during which the safety of the bridge with respect to the identified hazard will be assessed and the necessary mitigation measures will be designed. However, the current procedure shows some potential limitations. The calculation of the three main components that define the first risk estimate, the Class of Attention ( CoA ), composed of susceptibility, vulnerability, and exposure, does not account for the possible interaction between the landslide and the structure. Such interaction may indeed be inferred from signs of distress observable either on the bridge itself or in the surrounding area. As for susceptibility, the Guidelines require a preliminary and often simplified estimation of landslide characteristics such as volume, velocity, and state of activity, yet they do not include a specific assessment of how the landslide could interact with the structure. This interaction may manifest in very different ways depending both on the type of landslide and on the structural typology of the bridge (Gabrieli et al., 2024). In the current procedure, the type of interaction is considered only through three interference classes (total, partial, and approach zone) which are used in the calculation of vulnerability. This parameter plays a secondary role and merely allows the overall vulnerability level to be increased or decreased by one class. Another limitation of the landslide risk assessment concerns the small number of attention classes available, which are only five. As a result, this categorization does not allow for a true prioritization ranking among the assessed structures. When a large number of bridges fall into the same class, it becomes difficult to allocate available resources efficiently to those requiring more urgent and detailed investigations. In the inventory phase, this issue becomes particularly evident, since no real concept of prioritization is defined at this stage. As a result, there is no method available to assist in deciding which structure should be inspected before another. To address these prioritization challenges and to allow the integration of defects identified during inspection, this article proposes a method called the Prioritization Index ( PI ). The method consists of two coefficients: one that can be computed after the inventory and one after the inspection. Its purpose is to rank the assessed bridges and viaducts according to the urgency associated with their landslide-related risk. The procedure for calculating these coefficients is inspired by the aforementioned “defect sheets” used in structural risk assessment. In those sheets, each defect observed during inspection is assigned a weight based on its type, along with a severity level and an indication of its spatial extent. For the pre-inspection index, the calculation is performed considering only the information available in existing documentation, primarily records concerning the history of the structure and cartographic sources such as hazard maps and landslide inventories. The post-inspection index, on the other hand, is defined based on the precursor signs that may appear either on the structure or in the surrounding area, correlating these observations with the potential type of interaction between the landslide and the structure. Both coefficients were applied to a set of structures for which landslide risk had been certified, allowing for the calibration of the weights and levels used in the calculation. This made it possible to obtain an initial prioritization of the analyzed structures. The weighting factors were defined based on the authors’ field experience and on the expected severity of the potential consequences associated with each defect or interaction indicator. Further refinement and optimization of the weight calibration are expected as the application dataset is expanded.