Needs-driven Heterogeneous Multi-Robot Cooperation in Rescue Missions

TL;DR

This paper presents a needs-driven multi-robot cooperation framework for rescue missions, demonstrating that heterogeneity among robots improves performance, robustness, and efficiency in urban search and rescue tasks.

Contribution

It introduces a novel needs-driven cooperation mechanism using Behavior Trees for heterogeneous robots, with theoretical proof and simulation validation of improved group utility and robustness.

Findings

Heterogeneous robot systems outperform homogeneous ones in group utility.

Needs-driven cooperation reduces uncertainty and enhances robustness.

Simulation results confirm improved performance in rescue scenarios.

Abstract

This paper focuses on the teaming aspects and the role of heterogeneity in a multi-robot system applied to robot-aided urban search and rescue (USAR) missions. We propose a needs-driven multi-robot cooperation mechanism represented through a Behavior Tree structure and evaluate the system's performance in terms of the group utility and energy cost to achieve the rescue mission in a limited time. From the theoretical analysis, we prove that the needs-driven cooperation in a heterogeneous robot system enables higher group utility than a homogeneous robot system. We also perform simulation experiments to verify the proposed needs-driven collaboration and show that the heterogeneous multi-robot cooperation can achieve better performance and increase system robustness by reducing uncertainty in task execution. Finally, we discuss the application to human-robot teaming.