Distributed Multi Agent SLAM with Edge Intelligence and Ultra Low Latency Communication for Autonomous Systems

Abstract

Distributed multi-agent SLAM has emerged as a promising paradigm for enabling scalable and efficient autonomous navigation in complex environments. This study proposes a novel framework that integrates edge intelligence with ultra-low latency communication to enhance collaborative SLAM performance among multiple agents. The architecture leverages edge computing to perform localized data processing and feature extraction, significantly reducing computational load on individual robots while enabling real-time decision-making. Simultaneously, an ultra-low latency communication layer facilitates rapid sharing of keyframes, feature descriptors, and map updates across agents, ensuring synchronized mapping and cooperative loop closure detection. The distributed design improves robustness, scalability, and fault tolerance, allowing the system to maintain performance even in the presence of communication disruptions or agent failures. Experimental evaluations demonstrate improved localization accuracy, reduced trajectory drift, and enhanced communication efficiency compared to centralized and single-agent SLAM systems. The proposed framework offers a practical and scalable solution for next-generation autonomous systems operating in dynamic and resource constrained environments.