Collaborative Path Planning for Swarm UGV Operations in Dynamic Environments

Abstract

The use of swarm Unmanned Ground Vehicles (UGVs) offers significant advantages in a variety of applications, such as surveillance, exploration, and search-and-rescue missions. However, coordinating multiple UGVs in dynamic and unpredictable environments poses substantial challenges, particularly in path planning and collision avoidance. This paper presents an advanced framework for collaborative path planning in swarm UGV operations, addressing real-time communication, dynamic obstacle avoidance, and adaptive decision-making. Our approach leverages decentralized control strategies, where each UGV operates autonomously yet shares critical information with its peers to ensure effective swarm coordination. We integrate machine learning techniques for real-time path optimization and multi-agent systems (MAS) theory to enable collaboration among UGVs. Simulation and field tests demonstrate that the proposed framework significantly improves navigation efficiency, reduces collision risks, and enhances mission success rates, even in highly dynamic environments. The results indicate that this collaborative path planning approach is scalable and adaptable, providing robust performance across diverse operational scenarios.