Quantifying Resilience in Complex Systems: From Robustness to Adaptivity

Abstract

Real-world systems are inherently adaptive, capable of recovering and reorganising after disruptions. To truly understand and enhance resilience, we must quantify such adaptivity. Achieving this requires high-resolution, real-time data and mathematical models that capture not only the impact of disruptions but also the system’s capacity to respond and change. By integrating agent-based models with network analytics, we can track disruptions as they occur, predict failure cascades, and design more effective recovery strategies. This approach is especially crucial for infrastructures like drug supply chains, where real-time adaptation is key to mitigating shortages and ensuring the continuous delivery of pharmaceuticals to patients. In this talk, we demonstrate how moving beyond static resilience measures allows us to develop more comprehensive metrics that capture a system’s robustness and adaptive capacity, providing deeper insights into how they withstand and recover from shocks.