IEEE TRANSACTIONS ON SYSTEMS, MAN, AND CYBERNETICS—PART A: SYSTEMS AND HUMANS 1 A Socio-Physical Approach to Systemic Risk Reduction in Emergency Response and Preparedness William Ross, Student Member, IEEE, Alex Gorod, Member, IEEE, and Mihaela Ulieru, Member, IEEE Abstract—This paper proposes a socio-physical approach that considers jointly the interaction and integration of the social and physical views of a system to improve emergency response and preparedness. Using network analysis, it is shown that the explicit socio-physical approach yields meaningful qualitative and quantitative differences when compared with approaches that focus on the social and physical views in isolation. The benefits of this proposed approach are illustrated on a case study using clustering analysis and a proof-of-concept simulation. This new approach leads to risk reduction by enabling a more informed and coordinated response strategy following an incident and a better identification of possible consequences and preparation strategies prior to an incident. Index Terms—risk reduction, socio-physical view, clustering coefficient, emergency response and preparedness, systemic risk, situational awareness. I. I NTRODUCTION Emergency-response efforts in major recent disasters such as Hurricane Katrina (2005), Deepwater Horizon (2010), and the Japanese earthquake and tsunami (2011) have revealed that the current uni-dimensional risk-reduction strategies are insufficient and that there is a need for a holistic systemic approach [1]–[4]. Traditionally, emergency-response activities, both nationally and internationally, have focused on managing consequences during the aftermath of disasters with insuf- ficient emphasis placed on developing strategies a priori to reduce risk and minimize damage. Globally, the number of disasters has been growing, particularly in the least-equipped areas, where emergency preparedness efforts are constrained by existing financial resources, among other factors [5], [6]. Besides the large-scale crises caused by natural disasters, “normal” accidents can also lead to widespread devastation— in particular circumstances that can trigger chain reactions, as observed in [7]. Crises may also stem from social, economic, and political consequences [6]. Regardless of cause, it is imperative that emergency responders take into consideration both the social and physical implications resulting from their actions, allowing important interdependencies to be accounted for before and after a disaster [2], [8]. Considering the social and physical dimensions in isolation leads to a partial view of the problem space and, subsequently, to a marginal assessment of systemic risk [9]–[14]. Since risk- reduction strategies are based implicitly on the view taken of an emergency situation [15], [16], this paper proposes an W. Ross and M. Ulieru are with the Adaptive Risk Management Lab, University of New Brunswick, Fredericton, NB, Canada e-mail: (william.ross@unb.ca, ulieru@unb.ca). Alex Gorod is with SystemicNet, LLC in New York, NY, USA e-mail: (alex gorod@yahoo.com). Manuscript received ; revised. encompassing socio-physical view, which considers jointly the interaction and integration of the social and physical views of a system. This combined view leads to enhanced awareness of how the system operates, increasing the potential for improved emergency response and preparedness in the face of systemic risk. There have been several attempts to define and measure systemic risk [2], [9], [17]–[20]. In fact, the term finds its origin in financial systems, where it refers to “the risk that the failure of one financial institution (as a bank) could cause other interconnected institutions to fail and harm the economy as a whole” (Merriam-Webster). The term has also gained in popularity following the financial crisis of 2008, and numerous quantitative and qualitative analyses, metrics, best practices, and lessons learned can be extracted from the financial domain [21]–[24]. In emergency response, a parallel is to consider different system components, where a failure in one component could result in a failure that impacts not only other components, but the whole system, as well (e.g., electrical power failure). Using an integrative view of the system, such as the socio-physical view, is instrumental in improving responders’ awareness of systemic risk and in allowing them to consider appropriate risk-reduction strate- gies that can leverage resources effectively to protect critical infrastructure and services. Networks and their interactions are frequently the cause of the cascading failures that “are the most common mechanism by which local risks can become systemic” [1], [2], [25]– [28]. For example, scale-free-type networks are in the power- law form and, independent of network scale, are considered to be resilient to random attacks, yet are very vulnerable to deliberate attacks [29], [30]. This is just one example of how the underlying properties inherent within network structures can result in different failures and underscores the importance of network measures in increasing responder awareness. Other measures can be used, as well. In emergency response, for instance, the clustering coefficient, together with connectivity, can inform responders of the structural type of a network being examined, its distribution patterns, and underlying behaviour [30]–[33]—all of which can prove invaluable when facing the need to make difficult decisions (e.g., limited resources). These measures can provide insight into how to influence the network to reduce possible risks, making the entire system more resilient. In order to objectively demonstrate the extensiveness of the explicit, combined socio-physical view in comparison to the social and physical views in isolation, the clustering coefficient is used as a “measure of local connections, or ‘cliquishness”’ [31], [34], and it is hypothesized that the different nodes