This thesis explores the development of a distributed fleet management system for mobile robots using ROS 2, focusing on efficient multi-robot navigation, task allocation, and coordination. The system employs a master-slave architecture where a central master node manages navigation graphs, computes Closed Postman Problem (CPP) routes, and dynamically assigns waypoints to slave robots based on edge occupancy. Slave robots execute navigation commands and communicate status updates, ensuring robust task execution across the fleet. Additionally, the work includes advanced map processing techniques, transforming occupancy grid maps into topological graphs through Voronoi skeletonization and Euclidean distance mapping. This process enhances navigation efficiency by extracting navigable pathways and overlaying topological graphs onto original maps. The integration of graph theory, clustering, and real-time communication demonstrates practical applications in autonomous robot fleets for logistics, surveillance, and other dynamic environments.