Design and Development of Decentralized Swarm Network with Self-Reacting Capability

This paper presents the design and development of a decentralized swarm network for autonomous agents, capable of real-time self-reacting behaviors within a dynamically changing environment. Unlike conventional systems relying on centralized controls, this study leverages ESP32 microcontrollers, Firebase for data management, and a robust mesh network architecture to achieve seamless, self-organizing communication across agent robots. Utilizing the A* algorithm and wheel odometry, each robot adapts to environmental changes and coordinates effectively to accomplish collective goals, demonstrating resilience in adverse scenarios such as node failures or connectivity disruptions. Our results reveal the network's ability to maintain robust operation, ensuring adaptability and continuity in tasks such as surveillance, agriculture, and search and rescue missions. The proposed system stands out by reducing setup costs, enhancing scalability, and minimizing human intervention, pushing the boundaries of swarm robotics applications. Proportion of task load evenly distributed 85% with standard deviation 10% across all nodes while Average latency per node 0.2 seconds for 10 nodes, scaling to 0.35 seconds for 50 nodes. The decentralized nature of the system ensures that it adapts quickly to node failures or disruptions, leading to minimal recovery time and high connection stability. By leveraging ESP32 microcontrollers and Firebase, our system offers a more efficient and scalable solution, which contrasts with more conventional, centralized approaches that typically involve higher costs and complex setup.