Page 20 - Read Online

Clédel et al. J Surveill Secur Saf 2020;1:119­39 I http://dx.doi.org/10.20517/jsss.2020.08 Page 121 of 139

many measures and definitions are derived from more original ones, so that those that share a common origin
also share many characteristics. The current paper aims at gathering and comparing metrics and definitions of
resilience so that common criteria and differentiation criteria emerge from them. This way, categories of met-
rics and definitions can be defined. To identify pertinent literature, online database searching was performed
on databases such as Web of Science and DBLP. Articles were filtered with the keyword “resilience” and a
set of other keywords, including “metrics”, “measure”, evaluation”, and “framework”. The most relevant were
selected on the basis of their titles, abstracts, and whether they applied to the field of engineering. A second
step in this research consisted in cross-referencing the sources of the previously selected articles in order to
determine the origins of the particularities of their definitions and metrics.

This paper is organized as follows. Section 2 provides a survey of definitions of resilience, from its original
definition in ecological system to recent definitions in networks and cyber-physical systems. Definitions are
classified according to the ideas they focus on. Because there are many definitions for resilience, the expected
attributes of a resilient system can slightly differ from one article to an other. Thus, a description of the various
attributesassociatedwithresilienceisgiveninSection3. Then,asurveyofdifferentmetricsusedtoevaluatethe
resilience of systems is provided in Section 4. Some metrics consist in measuring separately some attributes of
resilienceandthencombiningthem. Othersevaluateresiliencewithoutconsideringwhatthevariouscapacities
that compose resilience, and they measure the impact of harmful events that occurred on a system to assess the
levelofresilienceofthissystemfortheseevents. Allconsideredmetricsareclassifiedaccordingtotheattributes
theytakeintoaccount. Theresultsofthisclassificationaresummarizedinatableattheendofthesection. Since
resilience is a complex property, it may often be confused with other concepts and system properties. Section
5 provides results of some articles that compare resilience with other properties such as robustness and risk
assessment. Section 6 discusses the existing limitations and gaps in the described definitions and metrics.
Additionally, it provides the conclusion of this study.

2. RESILIENCE DEFINITIONS
The term “resilience” comes from the Latin word “resilire’,’ which has several interpretations such as “to re-
bound”, to “spring back”, or “to withdraw into oneself”. Even if the current meaning of “resilience” differs
slightly from its Latin origin and despite the diversity of definitions, most of them fit with at least one of these
antic meanings. The resilience perspective emerged in the 70s from ecology with the work by Holling [11] . A
few years later, the resilience concept began to influence other fields such as anthropology, sociology, or psy-
chology, as described in [12] , before it reached engineering sciences and, even more recently, into computer
science and information technologies.

The notion of resilience was first developed in some domains such as ecology with the work by Holling [11] .
Resilience of a population is defined as a system property where the system behavior is less important than
the system persistence. Thus, resilience is distinguished from stability. The author described it as the capacity
of a system to move from a stability domain into another one and put the emphasis on “a high capability of
absorbing periodic extremes of fluctuations”, the maintainability of “flexibility above all else”, and a capacity
to “restore its ability to respond to subsequent unpredictable environmental changes”. Historically, resilience
has also been developed in psychology and refers to the ability to recover from trauma and crisis [13] while
“childhood resilience is the phenomenon of positive adaptation despite significant life adversities” [14] .

2.1. A system property
Francis and Bekera [15] described resilience as a system property to endure undesired events in order to ensure
“the continuity of normal system function”. This ability corresponds to three system’s capacities: absorptive,
adaptive, and restorative capacities. It could be considered that this definition goes against the original concept
ofresiliencegivenbyHolling [11] asthecontinuityofnormalfunctioncanbeconsideredasasynonymofsystem