Assesing the Impact of Smart Initial Response Strategies on the Resilience in Complex Systems

Abstract

During an emergency, the initial response and especially the identification of the damage in a complex engineered system is critical for the maintenance of the entire systemﾒs performance and operation. Simultaneously, todayﾒs smart technology offers new solutions, like the monitoring and the early warning systems that can effectively enhance the response processes. The purpose of this work is to investigate and to practically estimate the impacts of different failure recognition and initial response strategies on the resilience of a complex system that undergoes specific disruptions.

The research is based on an agent-based modeling approach and the findings are extracted from multiple repetitive simulations of the examined models. A generic agent-based model of a complex system is developed, comprising two discrete but interdependent subsystems, which both sustain a disturbance. The disruptive events address the connectivity of each subsystem and comprise three scenarios of increasing magnitude and complexity. Under each one of the disruptive scenarios, the research investigates four different versions of initial response strategies that are characterized by increasing levels of preparedness and automation in the related processes, as the level of the entire systemﾒs smartness also increases. The examined strategy procedures, which incorporate diverse smart technology features, are based on the inspections executed by the subsystemsﾒ nodes to identify the failure i.e., the missing connection in their respective subnetworks and finally, to restore it. In each one of the response strategies, these checks procedurally range from consecutive to parallel order, and from a reactive to a more proactive approach.

The first findings verify that smart response strategies generally improve the systemﾒs resilience in terms of performance maintenance and recovery time, regardless of the various interdependencies between the disrupted subsystems. Particularly, the effectiveness of the smarter responses is higher when the disturbance severity is greater. The findings also suggest that the specific features of the examined strategies and their contribution to the systemﾒs resilience seem to be associated with the quality of the disruption.

This research offers a new methodological approach for the quantitative assessment of the diverse initial response strategies, based on their effects on the systemﾒs resilience that evidently supports the selection and the deployment of the appropriate procedures and tools. Although this work features only a few of the various aspects of smartness, the findings promote a more effective introduction and orchestration of smart technology solutions into existing complex systems.