COSMOA an ontology-centric multi-agent system for co-ordinating medical responses to large-scale disasters

In recent years planning for large-scale disasters has become increasingly important. The advent of global terrorism has added a further impetus to such preparations particularly in the planning of the provision of emergency medical services. Whilst joint emergency service training exercises and pre-planning have improved the situation it is clear that further progress can be made. An emergency plan, made in advance will be rigid, by its very nature and may not take into account the current operational conditions at the time. Gathering information from distributed sources, integrating it and then making planning decisions are difficult tasks but are required in order to make plans more dynamic. It is clear that such complexity makes it impossible for any single human or even a team to fulfil the roles adequately. In this paper we discuss how multi-agent systems can be employed to help address this issue. We describe COSMOA an ontology-centric multi-agent system that is aimed at supporting hospitals during the response to a largescale disaster by producing a web-based emergency plan. The system exploits the abilities multi-agent systems have demonstrated in processing distributed data and making appropriate recommendations. A prototype of COSMOA has been implemented and we discuss some of the preliminary results. We also consider future work in this area and how this system could be expanded to cover more roles in health-