PSI - Issue 84

N. Kheirkhahan et al. / Procedia Structural Integrity 84 (2026) 33–40 N. Kheirkhahan et al./ Structural Integrity Procedia 00 (2026) 000–000

34 2

1. Introduction As the backbone of infrastructure systems, road networks underpin key social and economic activities, facilitating daily commuting, tourism, and visitor access (Jiang and Li, 2025). Additionally, road networks play a pivotal role in emergency operations during disruptive events and natural disasters. Therefore, ensuring that infrastructure systems can withstand multiple hazards is crucial for maintaining safety, reliability, and functionality, while also reducing economic impacts resulting from disruptions (ISTAT, 2024; Piscitelli et al., 2019). Natural hazard events affecting road networks are becoming more frequent. However, existing resilience frameworks are often inadequate because they do not fully account for hazard, exposure, and topological importance of network elements (Vareias et al., 2019). Hazards are potential phenomena, activities, or conditions that can cause casualties, harm people, damage property, or disrupt livelihoods and services with a certain probability in each area and at any given time (UNDRR, 2020). Hazards can occur separately or in combination with each other. In this latter case, the combined effects may exceed the sum of the individual hazards' effects. For this reason, multi-hazard analysis should be incorporated into resilience assessment and management frameworks (Kappes et al., 2012). Nevertheless, many resilience assessments overlook frequent geological hazards (Bera et al., 2020). In network theory, some road segments, called critical edges, are essential for maintaining connectivity. These edges have no alternative routes, so if they are blocked or fail, parts of the road network can become disconnected. Identifying these critical segments is important for understanding and improving network resilience (Kheirkhahan et al., 2025). These segments are essential for daily mobility and the functioning of road networks. Moreover, damage to key edges can cascade through supply chains (Giuffrè et al., 2021; Mitoulis et al., 2022) and severely restrict access to critical infrastructures (Cavalieri et al., 2020), making functionality under multiple hazards crucial (Pan et al., 2021). Network resilience, i.e. the ability to maintain function under disturbance (Holling, 1973), requires integrating real world data with graph-theoretic approaches (Hassan et al., 2022; Freeborough et al., 2016). Effective resilience assessment also accounts for cascading disruptions from nearby failures (Argyroudis et al., 2015; Byun et al., 2022), as observed during the Central Italy earthquakes (Fiorentino et al., 2018). Prior frameworks have primarily focused on connectivity between critical node pairs in urban areas (Miano et al., 2024), leaving a gap in network-wide hazard resilience (Zeleke et al., 2026). This need for comprehensive resilience assessment is further highlighted by Directive (EU) 2022/2557, known as the Critical Entities Resilience (CER) Directive, which places disaster risk reduction at the core of policies for the protection of critical infrastructures (as well as the essential services they provide), calling for a systematic approach grounded in the preventive analysis of vulnerabilities and in the adoption of suitable strategies to increase the resilience. The present study addresses this issue by proposing a comprehensive framework that integrates Geographic Information System (GIS)-based mapping, multi-hazard analysis (Cappucci et al., 2024), and complex network analysis to identify and prioritize road segments exposed to earthquakes, landslides, floods, and tsunamis. The framework assesses the impact of failures on network functionality and accessibility to critical infrastructure, deriving a criticality score for each segment to support data-driven decision-making and optimize maintenance tasks in terms of budget allocation and scheduling. One of the key strengths of the framework is its reliance on open access data, ensuring transparency, reproducibility, and broad applicability. The proposed methodology is applied to the road network of the municipality of Messina (Italy), which is considered a pertinent case study due to its role in European corridors and its high seismic hazard (Cavalieri et al., 2020). Moreover, the proposed approach can represent a concrete contribution to the development of monitoring and resilience strategies for critical infrastructures, in line with European objectives outlined in the CER Directive. 2. Case Study: The Municipality of Messina This study focuses on the municipality of Messina, whose boundaries are defined by Italian National Institute of Statistics (ISTAT) administrative data (ISTAT, 2025). The municipality’s road network is derived from OpenStreetMap (OSM; HOT, 2025) and includes non-pedestrian highway types classified according to Italy’s official road hierarchy (D’Angelo et al., 2025). Messina’s road system is represented as a directed graph comprising 5,371 nodes and 7,242 edges, enriched with detailed OSM attributes such as road classifications (Watts et al., 1998), where intersections are modeled as nodes and road segments as edges. The topological properties of Messina road