Infrastructure in Assessing Disaster-Relief Agents in the RoboCupRescue Simulation

The RoboCupRescue Simulation project has been implemented as one of the responses to recent large-scale natural disasters. In particular, the project provides a platform for assessing disaster-relief agents and simulations. However, its research evolution is limited because all agents’ programs must be developed by each researcher and the experimental operations are complex. To address these problems, we propose a combination of an agent development framework and experiment management software in this study as infrastructures in assessing disaster-relief agents in the RoboCupRescue Simulation. We have provided those elements separately; however, it becomes possible to easily carry out experiments that have flexible configuration by combining two elements. In the evaluation, a combinatorial experiment as a case study confirms the effectiveness of the environment and shows that the environment can contribute to future disaster response research that utilizes a multi-agent simulation.