A Decentralised Multi-Agent System for Emergency Resource Allocation in Metropolitan Regions: (Demonstration)

Nowadays, emergency departments in many countries still rely on human operators to allocate rescue resources to emergency events. However, with the fast growth of population in metropolitan regions, the frequency of occurrence of emergency events has risen significantly, which imposes tremendous pressure over emergency event operators. Agent and multi-agent systems (MASs) provide potential solutions to address such complicated emergency response problem due to agents ability of autonomous reasoning, adaptive decision making, complex environment modeling and collective group formulation [5]. In recent years, a number of promising agent-based emergency response and simulation systems have been developed to support rescue operations for different emergency situations. Some of the important ones include FireGrid [4], RoboCup Rescue [7], DEFACTO [10], DrillSim [2] and ALADDIN [1]. Nevertheless, most of these systems are limited to simulate and study emergency crises, which do not provide concrete solutions about how to improve the effectiveness of response operations and optimise the resource allocation results. In this demonstration, we present a MAS for the simulation of emergency rescue response in metropolitan regions, which is implemented based on our previous work in [11]. Our MAS provides a decentralised resource allocation approach to deploy resources to multiple emergency events simultaneously on Google Maps.