Article Preview
TopIntroduction
Without limitation to the Great East Japan Earthquake in 2011 that caused a large-scale serious damage, many disasters, such as earthquakes and tsunami, are enhancing people’s awareness for disaster risk reduction. Accordingly, there are a lot of studies for disaster reduction and emergency management; in many of them, utilization of advanced information and communication technologies are critical (Bessis, Asimakopoulou, Norrington, Thomas, Varagamto, 2011).
In the event of a disaster, it is very important to evacuate people quickly to evacuation sites in order to minimize the number of victims of the disaster. For this purpose, there is a need for a method for quickly determining evacuation routes and guiding people in accordance with the status of the locations of evacuation sites, the location of impassable road section, and the congestion caused by vast number of evacuees. However, as we cannot predict which areas will be damaged in the event of a disaster, it is extremely difficult to determine evacuation routes in advance of the occurrence of a disaster. In addition, it is difficult to gather and to share information under poor communication environments in the event of a disaster. Although the ad hoc network system, Nervenet, to connect remotely located nodes or to get a connection to isolated areas caused by the infrastructure damages under disasters are developed (Inoue, Ohnishi, Peng, Li, & Owada, 2011), delays till the activation of the ad hoc network for information distribution after the occurrence of a disaster are inevitable; delay to determine evacuation routes cannot be avoided. Therefore, it is necessary to determine evacuation guidance routes in consideration of delays that occur in information distribution.
In this paper, we propose a method of determining evacuation guidance routes in consideration of delays of information distribution. The system based on the method gathers wide-area information including information about the road network, the capacities of individual roads, the locations of evacuation sites, and the number of evacuees immediately before the occurrence of a disaster. Furthermore, the system uses an ad hoc network such as Nervenet and gathers information such as the road conditions and the accurate number of evacuees, after the disaster occurs, a certain time passes, and the ad hoc network begins to work. Then, the system determines evacuation routes by using a proposed algorithm; and it sends the evacuation routes to smart phones of evacuees.
We evaluate through simulation relationships among the average time to completion of evacuation, delay time that get wide-area information, and the locations of impassable road sections destroyed by the disaster. Furthermore, we evaluate the performance the proposed method and demonstrate the effectiveness of the method.