PSI - Issue 84

Fulvio Busatta et al. / Procedia Structural Integrity 84 (2026) 797–804

798

1. Introduction The Guidelines on risk classification and management, safety assessment, and monitoring of existing bridges that were issued by the Italian Ministry of Infrastructures and Transport in 2020 (CSLP, 2020), provide a multi-level procedure, wherein specific provisions are given also for the use of continuous monitoring (Fig. 1). In more detail, for those existing bridges and viaducts where the Risk Classification was either evaluated as Medium-High or High at Level 2, continuous monitoring is considered as an effective tool for the surveillance of the structural system. Moreover, monitoring is supposed to provide a sound basis for model updating when the Risk Class is High, so that accurate safety assessment (Level 4) is required.

GEOLOCATION / INVENTORY

Level 0

SPECIAL INSPECTION: pre-stressed bridges with post-tensioned tendons, bridges located in areas with evidence of flooding, erosion and landslide phenomena, or recognised being at high hydrogeological risk, where interference with the bridge structure is expected

SURVEYS / VISUAL INSPECTION and DEFECT RATING

Level 1

RISK CLASSIFICATION

Level 2

Bridges with evident issues due to their structural type and/or material (bridges with from high to severe inherent structural fragility)

HIGH

MEDIUM-HIGH

MEDIUM

MEDIUM-LOW

LOW

Periodic Inspection

Periodic Inspection

Periodic Inspection

SURVEILLANCE AND MONITORING SYSTEM

Bridges for which structural assessment is mandatory according to the Building Code NTC 2018

Continuous Monitoring

Non-routine Ad-hoc Inspection

Structural model updating

Level 3

Level 4

PRELIMINARY ASSESSMENT

ACCURATE ASSESSMENT

Operativity (t ref = 30 years) Transitability 1 (use restrictions, t ref = 5 years) Transitability 2 (traffic load limit, t ref = 5 years)

For all Risk Classes: Maintenance Planning & Scheduling (BMS Bridge Management System)

RISK CLASSIFICATION UPDATING

STRUCTURAL ASSESSMENT ACCORDING TO THE BUILDING CODE PROGRESSIVE USE OF BIM

Critical bridge structures

Traffic load monitoring

Level 5

UPGRADING INTERVENTIONS

NETWORK RESILIENCE ANALYSIS

Fig. 1. Flow diagram of the multi-level procedure established by the Italian Guidelines on existing bridges.

Within the framework of the (CSLP, 2020) Guidelines, remarkable funding was allocated by the Italian Government on the existing national motorway and roadway infrastructure network to complement the post-pandemic National Recovery and Resilience Plan financed by the Next Generation EU Plan. As a result, a large number of existing bridges and viaducts have been recently equipped with dynamic monitoring systems for Structural Health Monitoring (SHM) purposes. Among the asset managers that benefitted from the funding mentioned above, ANAS S.p.A. has been managing and continuously monitoring several hundred bridges through a dedicated cloud platform that employs an industrial Python-based version of the P3P software (García-Macías et al., 2023). Within the ANAS bridge stock under surveillance, the investigated “ Torrente Reggia ” bridge is a 5-span prestressed-concrete (PSC) continuous girder roadway bridge with span sequence 45.0 m + 90.0 m + 3×75.0 m. The bridge is equipped with a permanent SHM system including 35 MEMS accelerometers and 3 thermocouples. The continuous monitoring program of the bridge has been ongoing since 31 May 2024. Within SHM frameworks based on Operational Modal Analysis (OMA), novelty analysis aims to detect structural anomalies by inspecting the time evolution of natural frequencies estimated from ambient excitation. Novelty detection algorithms generally require a training period and multivariate control charts are often used to effectively detect anomalies that may occur in the testing period. In particular, since modal estimates, such as natural frequencies and mode shapes, are related to the stiffness and mass properties of the structure, their variation in time can be related to the structural integrity within the framework of Statistical Pattern Recognition paradigm (Sohn et al., 2001). The