Road Network Resilience: How to Identify Critical Links Subject to Day-to-Day Disruptions

Gauthier, Pauline; Furno, Angelo; Faouzi, Nour‐Eddin El · 2018 · OpenAlex-citations

DOI: 10.1177/0361198118792115

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This paper addresses the challenge of identifying critical links in road networks that are susceptible to day-to-day disruptions, such as accidents or adverse weather. While resilience analysis is crucial for maintaining mobility and managing emergency services, existing methods often rely on static topological metrics that ignore dynamic traffic properties like congestion and demand. The authors aim to determine whether purely topological approaches are adequate for measuring resilience and to propose a more accurate methodology that accounts for the dynamic, spatio-temporal, and demand-dependent nature of road traffic. To achieve this, the authors develop a methodological approach based on resilience stress testing using a dynamic mesoscopic simulator. This simulator, based on the Lighthill-Whitham-Richards model, captures traffic dynamics including demand, congestion, and dynamic route assignment. The stress testing methodology involves simulating disruptive events by gradually reducing link capacities (at levels of 0%, 20%, 40%, 60%, and 80%) or increasing traffic demand. The impact of these disruptions is quantified using an "Importance" metric, which measures the increase in overall travel cost (travel time divided by distance) weighted by traffic demand. The authors aggregate these performance losses into a "Stress Test Criticality" (STC) metric to rank links. This simulation-based approach is compared against four variants of Betweenness Centrality (BC), a traditional topological metric, including unweighted and travel-time weighted versions. The study evaluates these methods on two case studies: a simple virtual network with eight nodes and nine links, and a real-world network in the Paris agglomeration (DIRIF) comprising 868 links. Results from the virtual network demonstrate that link rankings vary significantly depending on the metric used. For instance, while topological metrics identified link 5 as the most critical due to its high frequency in shortest paths, the stress testing approach identified link 9 as more critical because its disruption caused severe congestion with no alternative routes for high-demand traffic. The simulation-based method successfully captured dynamic phenomena, such as traffic rerouting and the impact of link length on travel time, which static topological metrics failed to account for. Furthermore, changing traffic demand levels in the simulation altered the criticality rankings, highlighting the demand-sensitivity of the proposed approach. The significance of this work lies in demonstrating that purely static topological metrics are often inaccurate for assessing road network resilience because they ignore traffic demand and network dynamics. The authors conclude that while stress testing provides highly accurate results by incorporating these dynamic factors, it is computationally intensive and may be prohibitive for large networks. Conversely, topological metrics are computationally efficient but lack precision. The study highlights the need to combine traditional traffic-agnostic topological analysis with demand-aware dynamic stress-testing techniques to effectively identify and rank critical links in road networks.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• naturalistic crash near crash