An Agent-Based System to Minimize Earthquake-Induced Damages

Over 2000 earthquakes happen every year in Japan, so that this country is often called earthquake-ridden (Government of Japan, 2006). There exists a serious problem to prevent the occurrence of earthquake-induced disasters, such as fire, short-circuits, gas leakage, etc. With the recent advent of nation-wide telecommunication networks, real-time earthquake information can be received at every household, and it can thus be utilized to control consumer electronics and reduce the risk of these earthquake-provoked disasters (Kueppers, 2002). Real-time information about the seismic activity in Japan is provided by the national earthquake early warning system operated by the Japan Meteorological Agency, JMA (Doi, 2002). By sensing primary waves, this system can notify people (e.g. via radio and TV channels, a mobile phone subscription service, etc.) several seconds before the earthquake devastating secondary waves hit a specific area. The Japan Electronics and Information Technology Industries Association (JEITA) has recently introduced an automatic consumer electronics control system (JEITA, 2005). When an earthquake early warning is received, this system provides services, such as activating alarms, stopping gas, opening doors, and the like. This system cannot, however, comply with the specific situation at each particular household, as circumstantial information about who live/stay in, where they are, what they currently do, etc. is not utilized by the system. In the presented study, we propose an agent-based system for the earthquake-induced disaster prevention. The system uses household-specific knowledge and can provide for generally a higher level of safety for the inhabitants than existing systems with similar goals do. The proposed system realizes a distributed architecture – a design solution making it quite reliable in (post-)earthquake conditions. There are specialized agents installed in different places (e.g. of a house or a public facility) and called “room agents,” which are autonomous, monitor various appliances and people in the rooms, and can control the equipment and electronics, and guide evacuation when an earthquake happens. For the control, countermeasure agents processing different types of rules are set up. During an earthquake, a countermeasure agent receives earthquake data and selects appropriate constraints, which are to be used by the room agents. As there can often be conflicts when