Assessing Robustness in Multimodal Transportation Systems: A case study in Lisbon

Abstract

This research paper aims to contribute to a more sustainable mobility solution by proposing and empirically testing methods to assess the robustness of a network of a multimodal transport system, looking at aspects in the network topology. We hypothesise that the appropriate multilayered and traffic-sensitive modelling of this network can help characterise robustness and further understand the integration of different modalities. We modelled and examined the Lisbon public transports as a multiplex network, allowing us to integrate various modalities and grasp its characteristics and robustness performance. To accomplish this, we evaluate how the network decreases performance by removing nodes and edges. Comparing different failure scenarios, we concluded that we should use slightly different methods for each goal and type of removal (edge vs node). Those who involve recalculating metrics are usually the most effective. However, we identified a specific context where, counter-intuitively, that is not the case. Proposing a novel normalised version of AUC, we were able to compare side by side the robustness of each modality layer, regardless of their size and the whole network. Testing targeting strategies, we also observed that robustness tests needed to remove about half the network nodes to leave all the remaining nodes completely disconnected, and node removals are more effective than eliminating edges in this topology. Lastly, we simulate cascading events such as the breakdown of an entire transportation line. The results of this study allow decision-makers to understand the topological vulnerabilities of a transportation network.