Advances in the Group Analytic Network Process (GANP) are discussed to support hazards planning and emergency management under incomplete information. A GANP multi-criteria Decision Support System (DSS) is put forth that uses quadratic mathematical programming and interval preference information. Civil defense and emergency managers use the proposed DSS in emergency exercise to select among evacuation and “shelter-in-place” alternatives. It is shown that the proposed GANP architecture can improve decision-making transparency, emergency management effectiveness, and user satisfaction.
Floods, earthquakes and other natural hazards have disrupted human civilization since ancient times, often with devastating consequences [1]. Group decision making is an important characteristic of modern emergency planning and management [2]. For example, in the context of refugee emergency response and contingency planning, the Office of the United Nations High Commissioner on Refugees (UNHCR) [3] promotes the use of “roundtable” meetings in order facilitate group decision making: “The views of one agency may differ from others, but this will often be to the advantage of the planning process since it provides a useful forum for all assumptions to be questioned and refined. The end product is thus more realistic”. Civil defense and emergency management group decisions share several unique characteristics. First, the group must often make many complex and multi-faceted decisions in a short period of time, thereby contributing to a high “decision load”. Second, these decisions may have potentially serious consequences. Maier [4] uses the term “decision quality” to describe the degree to which a wrong decision could lead to catastrophic results. Third, the group decision must often be made with incomplete information (both in terms of quantity and quality), particularly in the early stages of a disaster because of emergency management problems. Accordingly, the development and application of group multi-criteria Decision Support Systems [5], [6] and [7] could be extremely valuable to support emergency management decisions.
The Group Analytic Hierarchy Process (GAHP) [8], [9] and [10] and the Group Analytic Network Process (GANP) [11] have great potential for use in group emergency management decision making. For example, Levy [12] describes the use of ANP to improve flood hazard mitigation in the 1998 Yangtze river floods which caused over 3,000 fatalities and affected up to 250 million people. In the ANP, networks with dependence and feedback are constructed, and emergency management judgments are made (or measurements performed) on pairs of elements with respect to a controlling element. An absolute scale (a special instance of a ratio scale with a constant multiplier equal to one, invariant under the identity transformation) of emergency management priorities (values) is derived from sets of pairwise comparisons and ratings. The criteria are pairwise compared with respect to the goal, the sub-criteria with respect to their parent criterion, and the emergency management alternatives with respect to the last level of sub-criteria above them. The emergency management priorities are then synthesized throughout the network to yield the overall priorities for the alternatives.
The remainder of the paper is structured as follows. Section 2 describes current methods for GAHP/GANP under conditions of uncertainty. Advances in mathematical and computer modeling for emergency management GANP are put forth in Section 3. In Section 4, a real world emergency exercise is discussed, involving several emergency management and civil defense groups. In particular, a chemical spill is simulated in the city of Brandon, Manitoba. With application to evacuation and shelter-in-place decisions, it is shown that the proposed GANP technique improves emergency management effectiveness, decision transparency, and user satisfaction. Finally, Section 5 summarizes the main conclusions.
The economic and societal losses due to natural disasters, technologic emergencies, and global pandemics are increasing, requiring more effective group decision making under uncertainty. A GANP DSS for emergency management was put forth that used quadratic programming and interval information to cope with a severe lack of information. With application to an emergency exercise in Brandon, the system was shown to help emergency and civil defense agencies respond to a hazardous materials incident. Specifically, the optimal emergency management alternative is identified as a “shelter-in-place” followed by an evacuation to a nearby location. Future research goals include extending the GANP approach to a larger network structure, and carrying out sensitivity analyses.
