The RoboCupRescue -Simulation League- is an international testbed for the simulation of software agents and robots performing Urban Search And Rescue (USAR) missions.
The main purpose of the RoboCupRescue Simulation Project is to provide emergency decision support through the integration of disaster information, prediction, planning, and human interface. Heterogeneous intelligent agents conduct search and rescue activities in this virtual disaster world. This problem introduces researchers to advanced and interdisciplinary research themes. In AI/Robotics research, for example, behavior strategy (e.g. multi-agent planning, realtime/anytime planning, heterogeneity of agents, robust planning, mixed-initiative planning) is a challenging problem.
The agent competition is a modular large-scale disaster simulation in real time, to which multiple agent teams can connect in order to reduce human casualties and damage to buildings. Typical agent types are police forces, ambulance teams, and fire brigades. The rescue domain represents a real multi-agent scenario since most of the encountered problems cannot be solved by a single agent. For example, fire brigades depend on police forces to clear blocked roads in order to extinguish fires. Therefore, team cooperation and coordination is highly required in this domain. Moreover, the task is challenging due to the limited communication bandwidth, the agent's limited perception and the difficulty of predicting how disasters evolve over time.