Visual Analytics Approach to User-Controlled Evacuation Scheduling∗

Application of the ideas of visual analytics is a promising approach to supporting decision making, in particular, where the problems have geographic (or spatial) and temporal aspects. Visual analytics may be especially helpful in time-critical applications, which pose hard challenges to decision support. We have designed a suite of tools to support transportation-planning tasks such as emergency evacuation of people from a disaster- affected area. The suite combines a tool for automated scheduling based on a genetic algorithm with visual analytics techniques allowing the user to evaluate tool results and direct its work. A transportation schedule, which is generated by the tool, is a complex construct involving geographical space, time, and heterogeneous objects (people and vehicles) with states and positions varying in time. We apply task-analytical approach to design techniques that could effectively support a human planner in the analysis of this complex information H. 1.2 [User/Machine Systems]: Human information processing - Visual Analytics; 1.6.9 [Visualization]: information visualization.