Date of Award

Spring 2015

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Aeronautics and Astronautics

First Advisor

Karen Marias

Committee Chair

Karen Marias

Committee Member 1

Daniel DeLaurentis

Committee Member 2

William Crossley

Committee Member 3

Abhijit Deshmukh

Abstract

Resilience is the ability to withstand and recover rapidly from disruptions. While this attribute has been the focus of research in several fields, in the case of system-of-systems (SoSs), addressing resilience is particularly interesting and challenging. As infrastructure SoSs, such as power, transportation, and communication networks, grow in complexity and interconnectivity, measuring and improving the resilience of these SoSs is vital in terms of safety and providing uninterrupted services. ^ The characteristics of systems-of-systems make analysis and design of resilience challenging. However, these features also offer opportunities to make SoSs resilient using unconventional methods. In this research, we present a new approach to the process of resilience design. The core idea behind the proposed design process is a set of system importance measures (SIMs) that identify systems crucial to overall resilience. Using the results from the SIMs, we determine appropriate strategies from a list of design principles to improve SoS resilience. The main contribution of this research is the development of an aid to design that provides specific guidance on where and how resources need to be targeted. Based on the needs of an SoS, decision-makers can iterate through the design process to identify a set of practical and effective design improvements. ^ We use two case studies to demonstrate how the SIM-based design process can inform decision-making in the context of SoS resilience. The first case study focuses on a naval warfare SoS and describes how the resilience framework can leverage existing simulation models to support end-to-end design. We proceed through stages of the design approach using an agent-based model (ABM) that enables us to demonstrate how simulation tools and analytical models help determine the necessary inputs for the design process and, subsequently, inform decision-making regarding SoS resilience. ^ The second case study considers the urban transportation network in Boston. This case study focuses on interpreting the results of the resilience framework and on describing how they can be used to guide design choices in large infrastructure networks. We use different resilience maps to highlight the range of design-related information that can be obtained from the framework. ^ Specific advantages of the SIM-based resilience design include: (1) incorporates SoS- specific features within existing risk-based design processes - the SIMs determine the relative importance of different systems based on their impacts on SoS-level performance, and suggestions for resilience improvement draw from design options that leverage SoS- specific characteristics, such as the ability to adapt quickly (such as add new systems or re-task existing ones) and to provide partial recovery of performance in the aftermath of a disruption; (2) allows rapid understanding of different areas of concern within the SoS - the visual nature of the resilience map (a key outcome of the SIM analysis) provides a useful way to summarize the current resilience of the SoS as well as point to key systems of concern; and (3) provides a platform for multiple analysts and decision- makers to study, modify, discuss and documentoptions for SoS.

Share

COinS