Identifying early-warning indicators of onset of sudden collapse in networked infrastructure systems against sequential disruptions

Abstract

Tsunamis, power blackouts, and distribution systems failure drastically affect the networked infrastructure systems which further affect a countries economy. Moreover, if these systems reach critical thresholds, they may experience disproportionate losses in the system�s functionality. Here we propose an approach to identify the critical thresholds and observe the presence of warning regions for real-world transportation systems. While attack tolerance of networked systems has been intensively studied for the disruptions originating from a single point of failure, there have been instances where real-world systems are subject to concurrent disruptions at multiple locations. We determine the entire robustness characteristics of transportation networks of disparate architecture subject to disruptions of varying sizes. Using United States Airspace Airport network and Indian Railways Network data, and synthetic networks as prototype class of systems, we study their responses to synthetic attack strategies of varying sizes. We also observe the significant relationships between network robustness and size of simultaneous disruptions for the complex networked infrastructures for random failures and targeted attacks. Our approach can serve as a paradigm to understand the point of sudden collapse in real-world systems, and the principle can be extended to other network infrastructures to address critical issues of risk management, resilience, and system safety.