Using functional decomposition to bridge the design gap between desired emergent multi‐agent‐system resilience and individual agent design

Abstract

Increasing the resilience of modern infrastructure systems is recognized as a priority by both the International Council on Systems Engineering and the National Academy of Engineering. Resilience answers the key stakeholder need for a stable and predictable system by withstanding, adapting to, and recovering from unexpected faults. Increasing resilience in multi‐agent systems is especially challenging because resilience is an emergent system‐level property rather than the sum of individual agent functions. This paper uses biological systems as a source of inspiration for resilient functions, examining the central question How can biologically inspired design be used to increase the emergent property of resilience in multi‐agent systems? The paper uses functional decomposition to break down the individual functions that result in resilience and transfer the properties to generalized systems. Accordingly, the central hypothesis examined in this article is If functional decomposition is performed on eusocial insect colonies, then generalizable approaches to increase the emergent property of multi‐agent system resilience can be identified. The results provide two contributions. The first contribution is the identification of six general functions based on eusocial insect behavior that influence resilience. The second contribution is a description of the process of identifying and transferring insect behaviors into generalized design‐for‐resilience guidance. To support these contributions, a case study applies biologically inspired functions to an emergency power service system and proposes tactics for the power system to improve its resilience. Thus, this article provides a key step towards our goal of using biologically inspired design to influence the emergent property of resilience in multi‐agent systems.