Abstract
A significant association among several robots in a multi-robot system (MRS) is essential for handling complicated tasks in dynamic situations while considering different physical limitations. This research explores and improves AI-driven team strategies for MRSs using the RoboCup Rescue Simulation (RCRS) testbed, which supports programmable rescue agents in disaster mitigation. We explore a cooperative strategy designed for the RCRS challenge that integrates a task modelling framework into the simulation to improve task distribution and handle cooperation requests from platoon agents. Furthermore, a communication protocol was created so that agents could exchange and use information quickly and effectively in various circumstances. This study showed that, when compared to the default agents, overall simulation performance was improved in complex map contexts. The techniques developed in this study can improve cooperative efforts in practical settings by filling in the gaps between workable MAS collaboration mechanisms and practical limitations for physical multi-robot systems.