PSI - Issue 84

Antonio Fiorentino et al. / Procedia Structural Integrity 84 (2026) 757–764

760

interact with the structure either directly or indirectly. A recurring operational challenge relates to the management of suspected or potential landslides, i.e., phenomena that are not formally inventoried but can be recognized during field surveys through morphological evidence and precursory signs (e.g., bulging, surface cracking) or, more generally, through geomorphological and stratigraphic settings predisposing to instability (e.g., toe erosion of an embankment or slope) (Fig. 2). In such cases, the absence of a formal inventory and the high uncertainty affecting key parameters of the phenomenon (magnitude/volume, activity state, kinematics) make it more difficult to define activation criteria and the scope of Special Inspections. This requires an extended interpretation of the context (diagnostic area, significant geomorphological area, and area of influence) and a cautious, well-motivated estimation of the parameters relevant for risk assessment (Operational Instruction 4.4.2.5b) (Perilli et Al., 2026). Finally, Level 4 requires a structured Knowledge Project, but cost and complexity demand optimized investigations: preliminary models after Special Inspections are strategic to coordinate campaigns, avoid redundancies, and integrate monitoring from the design stage to support decisions and possible updates of the landslide CoA.

Fig. 2. Classification of recognized landslides (Pa) and potential landslides (Pc) according to the national Guidelines (LLGG), based on activity state and/or degree of criticality (after DM 204/2022).

3. MARIE research project Within the methodological framework defined and consolidated through the activities of the FABRE Consortium, the MARIE project (Methodological Approaches for Risk Assessment in the Framework of Landslide–Bridge Interaction) is positioned as a natural evolution in both application and research. It addresses the operational difficulties identified (documentary gaps, uncertainties in the execution of Special Inspections, and limited standardization of knowledge content) and aims to make knowledge building in landslide–bridge interaction scenarios more coherent, reproducible, and robust, extending through to advanced modelling stages. The project starts from the systematization of case studies and interaction typologies, supporting a more effective application of the Guidelines and reducing interpretative ambiguities, also through proposals for operational instructions and priorities for future updates. In this context, MARIE devotes particular effort to a more specific and operational definition of Special Inspections as the technical junction between screening levels and accurate safety assessments, formalizing univocal activation criteria and a coherent set of activities tailored to different types of phenomena (slow-moving landslides and DSGSDs, rapid flows, rockfalls, slides). Special Inspections are therefore not conceived as generic “additional” investigations, but as a targeted phase aimed at: (i) identifying existing or potential distress/defect conditions attributable to slope instability; (ii) reliably estimating the potential impact of rapid kinematic processes; and (iii) supporting a reasoned decision between continuing within screening levels or proceeding to Level 4 accurate safety assessment. 3.1. Methodology The methodology developed within the MARIE project is based on a progressive, hierarchical, and knowledge driven approach, conceived to support the transition from risk classification stages to accurate safety assessments in cases of interaction between bridges and landslide phenomena. The MARIE method is articulated as a logical sequence of interconnected phases, starting from the collection and systematization of representative case studies and leading to the definition of a taxonomy of interaction scenarios (Fig. 3). This taxonomy is grounded in statistical evidence derived from national and international case studies, which identify landslide type, volume, velocity/activity state, and mode of interference as discriminant parameters for evaluating the effects on bridges. The analysis highlights a predominance of slow-moving phenomena, while also indicating a significant presence of rapid events with potentially more severe impacts (Gabrieli et al., 2024; Salciarini et al., 2024; Simeone et al., 2024; Scala et al., 2025).