Beyond traditional metrics: redefining urban metro network vulnerability with redundancy assessment

Previous studies have predominantly analyzed Urban Metro Network (UMN) vulnerability from topological and functional perspectives, often neglecting the impact of disruptions on alternative route availability. This research introduces a novel redundancy-based vulnerability analysis, assessing the reduction in travel alternatives following disruptions. The Montreal UMN is used as a case study, utilizing General Transit Feed Specification (GTFS) data from the Montreal Transit Authority and trip data from the 2018 Montreal Origin-Destination survey. Using the open-source platform Transition, we simulate shortest transit routes for each trip, generate alternative routes, and compute travel times. We define one targeted and three random failure scenarios, selected from 100 simulations, to evaluate network vulnerability to various disruption types. Indicators are formulated, calculated, and compared across all scenarios. Each failure scenario involves a sequence of consecutive metro station disruptions, leading to complete network shutdown. Findings reveal that the metro network is significantly more vulnerable to targeted disruptions than random ones. Among all indicators, functional ones related to users' travel time show greater sensitivity to disruption type, be it targeted or random. Vulnerability indicators exhibit the most substantial changes during initial disruptions, highlighting their critical impact. Although traditional approaches (topological and functional) show a direct relationship between the number of disruptions and changes in vulnerability indicators, this is not true for the redundancy-based vulnerability indicator. In this case, the primary determinants are the locations of disrupted stations and the network's geometry, rather than the number of disruptions.