Enhancing Infrastructure Resilience Under Conditions of Incomplete Knowledge of Interdependencies

Today’s infrastructures — such as road, rail, gas, electricity and ICT — are highly interdependent, and may best be viewed as multi-infrastructure systems. A key challenge in seeking to enhance the resilience of multi-infrastructure systems in practice relates to the fact that many interdependencies may be unknown to the operators of these infrastructures. How can we foster infrastructure resilience lacking complete knowledge of interdependencies? In addressing this question, we conceptualize the situation of a hypothetical infrastructure operator faced with incomplete knowledge of the interdependencies to which his infrastructure is exposed. Using a computer model which explicitly represents failure propagations and cascades within a multi-infrastructure system, we seek to identify robust investment strategies on the part of the operator to enhance infrastructure resilience. Our results show that a strategy of constructing redundant interdependencies may be the most robust option for a financially constrained infrastructure operator. These results are specific to the infrastructure configuration tested. However, the developed model may be tailored to the conditions of real-world infrastructure operators faced with a similar dilemma, ultimately helping to foster resilient infrastructures in an uncertain world.