Article Preview
Top1. Introduction
The mutual exchange of services of various interconnected and interdependent critical infrastructures, such as healthcare, energy, transport, manufacturing, financial, etc., is essential to the functioning and well-being of modern society (Maier, 2009; Mattsson & Jenelius, 2015). These services must be maintained to a satisfactory level even under threats and disruptive events. The latter becomes significantly more difficult when considering the interconnected nature of critical infrastructures. Failure to provide services can lead to domino and cascade effect that impacts on other infrastructures related to the initially affected one (Ouyang et al., 2012). Interconnected systems are therefore by definition less resilient to disruptions (Johanson, 2010). This thesis is confirmed, for instance, by the largest blackout in history that affected more than 600 million people in India in July 2012. Through a cascade effect, several other systems (transport, telecommunications, finance) also failed (Lai et al., 2013; Romero, 2012). Other examples are the growing number of hurricanes such as Sandy, Isabel, Harvey and Irma, that provoked not only human and material damage but also economic and production/service capacity failures (Saleh et al., 2017).
The notion of resilience is here-defined as “the capacity of a system(s) to recover, in a minimum time, with minimum costs (financial, human, workload, etc.) a certain functioning capacity on all dimensions of its performances”. Note that resilience is related to functioning and can be assessed by analyzing the functioning of systems during the following situations: (a) before a disruptive event, (b) during a disruptive event and (c) after a disruptive event. Improving systems resilience before any disruption occurs can reassure society’s vital needs.
Current resilience assessment approaches are oriented towards individual systems (Pursiainen, 2018), whether they be financial, healthcare or transport systems. These approaches are therefore inflexible (difficult to adapt to other domains), with fixed criteria, mainly performance (other criteria that might be important in the assessment of resilience are overlooked) and not applicable in the context of interconnected systems. This problem is addressed by methods specifically designed for interconnected systems. For instance Kamissoko et al. (2018) introduce a method for “continuous and multidimensional assessment of resilience based on functionality analysis for interconnected systems”. This method provides the means to define functional analysis of interconnected systems, that is linked to some aspects of resilience. The functional analysis is further used to assess resilience, based on several criteria. The proposed generic criteria are easily extendable and adaptable depending on the context and needs. The originality of the method is defined by the authors as: (1) the combination of functional-analysis and continuous resilience assessment following several dimensions of critical infrastructures, (2) the flexibility of criteria metrics for easy adaptation in different contexts and (3) the possibility to aggregate the results of several functionality-analysis models with continuous assessment of the resilience of interconnected systems. The limit of this proposal, however, is in the modeling of interconnected systems and functional analysis. Several core concepts related to the modeling of complex systems (ISO/IEC 15288, 2008) are overlooked (e.g., physical and contextual modeling, influence and dependency modeling, behavioral and functional architecture, etc.). Indeed, an improved modeling (1) eases engineers’ design and analysis work, (2) provides to stakeholders a better understanding of systems, and (3) helps stakeholders to make decisions with confidence and to argument decisions based on both, models and analyses.