The efficient and emergent rescue operations are critical in a large disaster. However, the locations of demanders (those who need special care or help) and responders (those who care or help demanders) are often separated in space and apart in time. In this paper, we propose a method by which we can efficiently dispatch each responder to demanders, and show the best travel route.
 The above problem can be considered as a multiple Traveling Salesman Problem (mTSP). Firstly, with reference to the past studies on mTSP, we proposed a method to support an efficient travel for rescue activities. A fuzzy c-means clustering and a genetic algorithm (GA) were employed to dispatch responders to demanders and to determine the best travel routes. Furthermore, assuming specific needs at the event of disaster, we constructed several advanced calculation methods as follows:
 (1) A method to take into consideration of the difference of workload of demanders, in which a large workload was temporarily divided into multiple small unit workloads before dispatching the responders.
 (2) A method to consider the adaptability of responders, and the emergency for demanders, in which we constructed a sort of distance measure, which is composed by three indices (D, R, and E from the viewpoint of distance, adaptability, and emergency, respectively). Although it is necessary to quantify the weight of each index, it will be possible to adjust various demanders and responders.
 Next, the effectiveness and efficiency of the proposed method was evaluated by simulation experiments using the actual street network data. We examined the case that 10 responders cared 50 demanders by considering the difference of the workload of demanders (300 seconds/person for 10 demanders and 10 seconds/person for 40 demanders). The results showed the proposed method was efficient on shortening the time until all responders completed travel. Furthermore, we demonstrated that it was better to immediately recalculate the dispatch of responders to demanders to reduce the total travel time, in case that new responders or demanders appeared after dispatch.
 In addition, field experiments were conducted to confirm the usefulness of the proposed method in actual urban spaces. Integrating the proposed method into a Web application developed in the previous studies, responders were adequately dispatched to demanders and efficiently navigated by the monitor screen of mobile phone. The comparative experiments were also conducted using an existing SNS (LINE). The results demonstrated that the effective and efficient travel for rescue activities avoiding oversight and duplication of demanders were achieved by using the Web application incorporated the proposed method.