On the resilience of systems of systems

The need to consider how systems can be made resilient to failure modes has gained increasing traction in the fields of systems thinking and systems design, and is now more widely studied, with authors identifying the potential disruptive effects of failure upon a system, and codifying these disruptions into specific types. When the focus of specification moves from the bounded single system to the consideration of capability and effect, systems-of-systems, rather than systems must be contended with. Systems-of-systems have been classified as being of a number of types (acknowledged, collaborative, directed, virtual, for example), whilst authors have endeavoured to characterise the properties of systems-of-systems, and the difficulties associated with their design, introduction and operation. This study has invariably arrived at the conclusion that systems-of-systems are infinitely more complex than bounded single systems, and as the final system-of-systems design will still need to be resilient to failure, this in turn poses more difficult questions for the study of resilience, as the properties of a bounded single system are unlikely to be the same as those of a system-of-systems. This paper will consider the problems faced by the need to specify resilience in a system-of-systems environment, by first evaluating how the various types and properties of systems-of-systems might affect the consideration of resilience, and then proposes an initial codification of systems-of-systems resilience disruption types, along with recommendations and required further work.