Model-driven decision support system for estimating number of ambulances required during earthquake disaster relief operation

Most of human life has been encountered danger due to natural disasters nowadays. One of these natural disasters that endanger human lives and which causes lot of damages is earthquake. A proper emergency response after an earthquake happening is important and has high priority in earthquake emergency management to reduce number of damages. Decision making for critical resources in the phase of response, is one of the main concerns for managers. Ambulance, as one of the critical resource that can help to reduce earthquake losses and costs, needs to be planned. Confusion in the number of victims in the early stages of earthquake, access complexity to the required data of different organizations by the pressing time, complicated nature of estimation, diversity of models and limitation of time for decision making are the main problems associated with estimating ambulances during earthquake disaster which makes estimation too difficult. In addition, there is a call for research in determining the number of required ambulances during earthquake emergency management, due to high error in estimating the number of ambulances in the current methods, which leads to unnecessary expenses and thereby helping to ensure that disaster sites are not overcrowded with emergency workers impeding each other's effectiveness. Such complexity suggests the introduction of Decision Support System (DSS). More accurate estimation of the number of required ambulances using a decision support system can help managers to speed up the process of decision making and thus reducing error and costs. Since the number of ambulances needed during a disaster is directly proportional to the number of victims requiring hospital treatment and in order to reach the first objective of this study, factors determining the number of human casualties in earthquake disaster i.e. population, modified Mercalli, age, time, building occupancy and gender are selected as the most relevant factors which have high probability in creating human casualties. The collected data from various relevant sources is used in proposing the model of this research. After testing different approaches, Fuzzy rule-based approach is being used, after defining the rules for each aforementioned factors and optimization is conducted in order to minimize the error for estimating the number of human casualties. Finally, by using de Boer formula and obtained number of human casualties, the number of required ambulances is estimated accurately. The results indicate that the error is decreased by more than 50% in the proposed method. A prototype of Model-Driven Decision Support System was developed based on the proposed model that can be used to aid emergency response planners for their decision making process prior to take any action during earthquake emergency management