Models of feedback and adaptation in multi-agent systems for disaster situation management

Abstract

The response, rescue and recovery teams that are engaged in disaster management operations require a continuous and comprehensive information flow of the disaster environment and a situational awareness in order to undertake fast and coordinated actions. Because of highly dynamic and often unpredictable disaster situations the teams need to adjust their goals, resources and actions both on an individual member level (agent) and on an entire team level (multi-agent system). This paper investigates a new approach to an agent's adaptability based on cognitive feedback introduced into the framework of inter-agent collaboration. The paper is a continuation of our work on situation-aware multi-agent systems. We discuss how agent adaptation and cognitive feedback is applied in the architecture of multi-agent systems for disaster situation management.