An Investigation of Ecologically-inspired Architecting Principles for Resilient System of Systems Design

Abstract

System of Systems (SoS) are networked integration of constituent systems that together achieve new capabilities not possible through the operation of any single system. Their resilience (being able to withstand and recover from disruptions) is a critical attribute whose evaluation is nontrivial and requires detailed disruption models. This forces a reliance on qualitative guidelines (such as redundancy, and localized capacity) for resilient SoS design. However, excessive reliance on these strategies can lead to unacceptable costs, waste or emissions, and the over-consumption of natural resources. Biological ecosystems are naturally existing resilient SoS. Ecologists have found that biological ecosystems achieve a simultaneously resilient and sustainable (efficient) design through a unique balance of constraints and redundancies in their network architectures. This architectural feature is measured using the metric Degree of System Order (DoSO). Motivated by this finding, the present research investigates the value of the DoSO metric and this biologically-inspired architecting principle for the design of resilient SoS. This research is the first to provide quantitative evidence suggesting that the ecological principle of balancing pathway constraints and redundancies can guide the design of SoSs with desirable resilience and affordability trade-offs. This work also develops an extended modeling framework for the DoSO evaluation of Cyber Physical SoS, and SoS with inventory (storage) at the constituent systems. The extended framework is tested on a power grid and a supply chain case study, finding promising indications that the DoSO analysis can evaluate SoS architectures’ fitness for resilience. The proposed approach is also validated against an external microgrid resilience analysis tool. Finally, this work also investigates the ecological principle of nestedness: an architectural characteristic that is prevalent in mutualistic biological ecosystems and has been associated with effective resource utilization, and the ability to avoid cascading extinctions. Through an investigation of industrial symbiosis networks, this research provides evidence that nested architectures can significantly reduce resource consumption, and provide lower operating costs under specific operating conditions in sustainability-focused SoS.