A Risk-Science Approach to Vulnerability Classification

Advancements in the risk literature and recent events have highlighted the need for recognizing and managing system vulnerabilities. However, established definitions of vulnerability typically involve only static concepts that are limited to measurement of system characteristics. Advancements in risk modeling, combined with the dynamic nature of data availability, and processing call for the need to understand the various dimensions and time-dependent properties of vulnerability within risk-informed decision making. There is need to: (1) Understand and classify aspects of vulnerability that exist in various systems, such as related to engineering, business, and healthcare, while recognizing both properties of the system and associated knowledge, (2) reconcile these definitions of vulnerabilities with existing concepts, such as sensitivity analysis and fragility, and (3) explore the implications of various types of vulnerability on risk management decisions. The main contributions of this work include classifying dynamic characteristics of system vulnerability and leveraging information about the multidimensional properties of vulnerability within risk management decisions that apply to a collection of risk events. As a proof of concept, we illustrate the vulnerability classification on the COVID-19 pandemic. This article will be of interest to both risk researchers and practitioners.