A multi-agent evacuation simulator to design safe cities for high quality of life with computer clustering

To achieve a high quality of life, it is necessary to live in environment which is safe from disasters. This paper deals with a simulation system that is utilized for designing a large scale environment such as city. Gigantic number of autonomous agents which evacuate from a disaster in a modelled city are simulated in it. In spite of simulating a large number of agents and wide area of environment, the simulator also computes very detailed local areas precisely. Generally, larger scale simulation with smaller granularity requires more computing performance. The authors employed a computer clustering technology to overcome this difficulty. The basic design are presented in this paper.