DEVELOPMENT OF A HYBRID COORDINATION ARCHITECTURE FOR A MOBILE ROBOT FLEET BASED ON CENTRALIZED PLANNING AND DECENTRALIZED NAVIGATION
https://doi.org/10.53360/2788-7995-2026-3(23)-9
Abstract
This paper proposes and formalizes a hybrid architecture for controlling a fleet of mobile robots. The centralized MAPF/CBS layer generates a consistent discrete motion plan, while the decentralized ROS2/Nav2 layer ensures continuous execution and local adaptation in a Gazebo environment. Unlike fully centralized schemes, which are sensitive to increasing computational complexity and execution deviations, as well as the Nav2-only mode, which lacks a global mechanism for preventing inter-robot conflicts, the proposed architecture introduces an intermediate Fleet Manager. This module converts the discrete plan into a sequence of waypoint commands, monitors progress, detects delays, and initiates replanning when motion conditions are violated. A mathematical formulation of the problem on a graph G=(V,E) is formulated, conflict constraints on vertices and edges, makespan and sum-of-costs target criteria are specified, and an Ackermann-type kinematic model is presented for continuous execution. Experimental verification was performed at two levels: in a discrete MAPF/CBS model and in the ROS2/Gazebo/Nav2 simulation environment. For the R1 reference mode, it was shown that as the number of agents on map A increases from N=1 to N=8, the planning time increases from 0.095823 to 15.709505 s, the SOC increases from 3 to 40, and the makespan increases from 2 to 4 cycles. The obtained results confirm the need for a hybrid transfer of a centralized plan to a locally adaptive implementation and define a reproducible framework for assessing the scalability, robustness, and conflict tolerance of a multi-robot system.
About the Authors
E. A. TuleshovKazakhstan
Yerkebulan Tuleshov – Candidate of Technical Sciences, Associate Professor of the Department of Robotics and Technical Means of Automation
050000, Republic of Kazakhstan, Almaty, Satpayev St. 22
K. M. Abutalipov
Kazakhstan
Kaisar Abutalipov – masters of Department of Robotics and Technical Means of Automation
050000, Republic of Kazakhstan, Almaty, Satpayev St. 22
M. Zh. Kuatova
Kazakhstan
Moldir Zhangeldievna Kuatova – PhD, Leading Researcher, RSE on the Right of Economic Management; Director of the Department of Scientific and Innovative Activities, International University of Engineering and Technology
050010, Republic of Kazakhstan, Almaty, Shevchenko St. 28
P. M. Rakhmetova
Kazakhstan
Perizat Rakhmetova – PhD, Associate Professor of the Department of Robotics and technical means of automation
050000, Republic of Kazakhstan, Almaty, Satpayev St. 22
References
1. Dai Y. Development of a Fleet Management System for Multiple Robots’ Task Allocation Using Deep Reinforcement Learning / Y. Dai, D. Kim, K. Lee // Processes. – 2024. – № 12(12). – Р. 2921. https://doi.org/10.3390/pr12122921.
2. Efficient multi-robot path planning in real environments: a centralized coordination system / D.M. Matos et al // Int J Intell Robot Appl. – 2025. – № 9. – Р. 217-244. https://doi.org/10.1007/s41315-024-00378-3.
3. Decentralized multi-robot formation control in environments with non-convex and dynamic obstacles based on path planning algorithms / L.E. Ruiz-Fernandez et al // Intel Serv Robotics. – 2025. – № 18. – Р. 215-232. https://doi.org/10.1007/s11370-024-00582-x.
4. Bayer J. Decentralized multi-robot exploration under low-bandwidth communications / J. Bayer, J. Faigl // Auton Robot. – 2026. – № 50. – Р. 7. https://doi.org/10.1007/s10514-025-10234-3.
5. Consistent batch fusion for decentralized multi-robot cooperative localization / N. Hao et al // Control Theory Technol. – 2024. – № 22. – Р. 638-651. https://doi.org/10.1007/s11768-024-00229-3.
6. Niu L. Decentralized Density Control of Multi-Robot Systems Using PDE-Constrained Optimization / L. Niu, G. Notomista // IEEE Robotics and Automation Letters. – 2025. – vol. 10, № 4. – Р. 4045-4052. https://doi.org/10.1109/LRA.2025.3548501. https://ieeexplore.ieee.org/document/10910164.
7. Centralization vs. decentralization in multi-robot sweep coverage with ground robots and UAVs / A. Jamshidpey et al // Artif Life Robotics. – 2025. https://doi.org/10.1007/s10015-025-01049-7.
8. Priority-Based Distributed Coordination for Heterogeneous Multi-Robot Systems With Realistic Assumptions / M. Cecchi et al // IEEE Robotics and Automation Letters. – 2021. – vol. 6, № 3. – Р. 6131-6138. https://doi.org/10.1109/LRA.2021.3091016.
9. Tahir N. Consensus-based Resource Scheduling for Collaborative Multi-Robot Tasks / N. Tahir, R. Parasuraman // Seventh IEEE International Conference on Robotic Computing (IRC), Laguna Hills, CA, USA. – 2023. – Р. 323-330. https://doi.org/10.1109/IRC59093.2023.00059. https://ieeexplore.ieee.org/document/10473544.
10. Comparative Analysis of Centralized and Federated Learning Techniques for Sensor Diagnosis Applied to Cooperative Localization for Multi-Robot Systems / Z. El Mawas et al // Sensors. – 2023. – № 23(17). – Р. 7351. https://doi.org/10.3390/s23177351.
Review
For citations:
Tuleshov E.A., Abutalipov K.M., Kuatova M.Zh., Rakhmetova P.M. DEVELOPMENT OF A HYBRID COORDINATION ARCHITECTURE FOR A MOBILE ROBOT FLEET BASED ON CENTRALIZED PLANNING AND DECENTRALIZED NAVIGATION. Bulletin of Shakarim University. Technical Sciences. 2026;(3(23)):83-90. (In Russ.) https://doi.org/10.53360/2788-7995-2026-3(23)-9
JATS XML















