Argonne National Laboratory Resilience Index

A very different approach to measuring the resilience of critical infrastructure is described by Fisher et al. (2010), the result of a project conducted by Argonne National Laboratory in collaboration with the U.S. Department of Homeland Security’s Protective Security Coordination Division. Data are gathered at critical infrastructure facilities by trained interviewers known as Protective Security Advisors (PSAs). The interviews use an Infrastructure Survey Tool covering roughly 1,500 variables that cover six major physical and human components (physical security, security management, security force, information sharing, protective measures assessment, and dependencies) that are themselves broken down into 42 components. The approach is used for one or several types of critical infrastructure or key resource sector (banking and finance, dams, energy, etc.). Data are subjected to an elaborate, six-step process of quality control involving review by experts in critical infrastructure protection.

A five-stage aggregation process is then used to combine the items into a single Resilience Index (called the Protective Measure Index PMI) that ranges from 0 (lowest resilience) to 100 (highest resilience) for a given critical infrastructure or key resource sector and for a given threat. Each of the stages takes a subset of items at that stage and combines them using weights to obtain a single index for the next stage. From roughly 1,500 items at Level 5, this process results in 47 composite scores at Level 2, three at Level 1, and finally a single score. At Level 2, 18 of the 47 measures contribute to Robustness at Level 1, five to Recovery at Level 1, and 24 to Resourcefulness at Level 1. At each stage, every contributing measure is multiplied by a weight, and the products are summed to obtain the PMI composite index. Weights are obtained by analyzing the opinions of experts, using the MCDM methods of Keeney and Raiffa (1976). PMI ratings by sector (e.g., commercial facilities, energy, transportation, water) may help in identifying the infrastructure facility that is weakest in relation to one or several threats.

In contrast to the bottom-up elements of the Coastal Resilience Index, this approach is almost entirely top down, reflecting the need of a national program to be uniform and universal in its approach. There is no possibility of adaptation to local needs, by modifying either the set of data items or the weights, both of which are prescribed. The index is entirely concerned with critical infrastructure, such a narrow focus being more conducive to a rigorous, quantitative approach. Nevertheless, justifying universal weights resolved to three decimal places is difficult given the inherent vagueness of the concept of resilience and its essential components, and uncertainties over the exact nature of threats.

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement