Skip to main content
Log in

3D multi-robot patrolling with a two-level coordination strategy

  • Published:
Autonomous Robots Aims and scope Submit manuscript


Teams of UGVs patrolling harsh and complex 3D environments can experience interference and spatial conflicts with one another. Neglecting the occurrence of these events crucially hinders both soundness and reliability of a patrolling process. This work presents a distributed multi-robot patrolling technique, which uses a two-level coordination strategy to minimize and explicitly manage the occurrence of conflicts and interference. The first level guides the agents to single out exclusive target nodes on a topological map. This target selection relies on a shared idleness representation and a coordination mechanism preventing topological conflicts. The second level hosts coordination strategies based on a metric representation of space and is supported by a 3D SLAM system. Here, each robot path planner negotiates spatial conflicts by applying a multi-robot traversability function. Continuous interactions between these two levels ensure coordination and conflicts resolution. Both simulations and real-world experiments are presented to validate the performances of the proposed patrolling strategy in 3D environments. Results show this is a promising solution for managing spatial conflicts and preventing deadlocks.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11
Fig. 12
Fig. 13
Fig. 14
Fig. 15
Fig. 16
Fig. 17
Fig. 18
Fig. 19
Fig. 20
Fig. 21
Fig. 22

Similar content being viewed by others



  2. This can be a rotating laser range-finder or a full 3D scanner.

  3. That is, the distance between the closest pair of points of the two planned paths is smaller than \(D_s\).

  4. We use an “is_” prefix to denote boolean variables.

  5. Or at a pre-fixed frequency, after a first selected is broadcast along the way to the current goal.

  6. In our case, this depends on the idlenesses of the nodes.

  7. The metric level modules must run on the robot main board and share computational resources with other demanding processing nodes (Kruijff-Korbayová et al. 2015).

  8. At this stage, we found this approach to perform very well in practice without significantly limiting the robot manoeuvres in the tested scenarios.

  9. Here we include the segmented obstacles in the map and the most recent nearby obstacle points which have been detected by the rangefinder and are not segmented yet in the map.

  10. The sub-optimality of the solution is due to the used incremental sampling-based approach (Karaman and Frazzoli 2010; Diankov and Kuffner 2007).

  11. As explained in Sect. 7.2, each point of \(\mathcal{M}_t\) can be associated to a robot pose.

  12. This can be used for instance to steer the robot toward regions where an estimated WIFI radio signal strength map returns higher values (Caccamo et al. 2017).

  13. The dynamic update of the OctoMap and its reactive behaviour is demonstrated in a video


  15. This aspect can be managed for instance as proposed in Zimmermann et al. (2014) and Colas et al. (2013).


  17. Two simulation videos are available on our website and show these behaviour.

  18. Since V-REP simulations are computationally demanding in our setup, the simulated robots were not able to move in real time and their motions were very slow (this can be observed in our simulation videos on our website). As a result, when robots got in interference, they persisted in such conditions for longer times with respect to a normal real time simulation.

  19. Which we do not report here in order to reduce space.

  20. This is not visible in the plot but it was observed by inspecting the recorded data.


  22. In these cases, the path planner only considers the most interesting and useful part of the traversability map.

  23. This laser proximity checker inhibits forward velocity commands when a close front obstacle is detected by the laser.

  24. In our setup, V-REP is not able to stably simulate more than four robots under realistic conditions (cfr. Sect. 10.1).

  25. The path and idleness message sizes actually depends on the number of patrolling graph nodes.

  26. Recurring to simpler and more affordable robotic platforms is required.



  • Acevedo, J. J., Arrue, B. C., Daz-Bez, J. M., Ventura, I., Maza, I., & Ollero, A. (2013). Decentralized strategy to ensure information propagation in area monitoring missions with a team of UAVs under limited communications. In 2013 International Conference on Unmanned Aircraft Systems (ICUAS) (pp. 565–574).

  • Acevedo, J. J., Arrue, B. C., Maza, I., & Ollero, A. (2016). A distributed algorithm for area partitioning in grid-shape and vector-shape configurations with multiple aerial robots. Journal of Intelligent & Robotic Systems, 84(1), 543–557.

    Article  Google Scholar 

  • Agmon, N., Kaminka, G. A., & Kraus, S. (2014). Multi-robot adversarial patrolling: Facing a full-knowledge opponent. CoRR abs/1401.3903.

  • Agmon, N., Kraus, S., & Kaminka, G. A. (2008a). Multi-robot perimeter patrol in adversarial settings. In ICRA (pp. 2339–2345).

  • Agmon, N., Sadov, V., Kaminka, G. A., & Kraus, S. (2008b). The impact of adversarial knowledge on adversarial planning in perimeter patrol. In Proceedings of the 7th international joint conference on autonomous agents and multiagent systems—Volume 1, AAMAS’08 (pp. 55–62). International Foundation for Autonomous Agents and Multiagent Systems.

  • Ahmadi, M., & Stone, P. (2006). A multi-robot system for continuous area sweeping tasks. In ICRA (pp. 1724–1729).

  • Aksaray, D., Leahy, K., & Belta, C. (2015). Distributed multi-agent persistent surveillance under temporal logic constraints. IFAC-PapersOnLine, 48(22), 174–179.

    Article  Google Scholar 

  • Andrade, R. D. C., Macedo, H. T., Ramalho, G. L., & Ferraz, C. A. (2001). Distributed mobile autonomous agents in network management. In Proceedings of international conference on parallel and distributed processing techniques and applications.

  • Baran, P. (1964). On distributed communications networks. IEEE Transactions on Communications Systems, 12(1), 1–9.

    Article  Google Scholar 

  • Barraquand, J., Langlois, B., & Latombe, J. C. (1992). Numerical potential field techniques for robot path planning. IEEE Transactions on Systems, Man, and Cybernetics, 22(2), 224–241.

    Article  MathSciNet  Google Scholar 

  • Bereg, S., Caraballo, L. E., Díaz-Báñez, J. M., & Lopez, M. A. (2016). Resilience of a synchronized multi-agent system. ArXiv e-prints.

  • Cabrita, G., Sousa, P., Marques, L., & De Almeida, A. (2010). Infrastructure monitoring with multi-robot teams. In IROS (pp. 18–22).

  • Caccamo, S., Parasuraman, R., Freda, L., Gianni, M., & Ögren, P. (2017). Rcamp: A resilient communication-aware motion planner for mobile robots with autonomous repair of wireless connectivity. In 2017 IEEE/RSJ international conference on intelligent robots and systems (IROS). IEEE.

  • Cadena, C., Carlone, L., Carrillo, H., Latif, Y., Scaramuzza, D., Neira, J., et al. (2016). Past, present, and future of simultaneous localization and mapping: Towards the robust-perception age. IEEE Transactions on Robotics, 32(6), 1309–1332.

    Article  Google Scholar 

  • Chen, H., Cheng, T., & Wise, S. (2017). Developing an online cooperative police patrol routing strategy. Computers, Environment and Urban Systems, 62, 19–29.

    Article  Google Scholar 

  • Chevaleyre, Y. (2004). Theoretical analysis of the multi-agent patrolling problem. In Proceedings of the IEEE/WIC/ACM international conference on intelligent agent technology (pp. 302–308).

  • Colas, F., Mahesh, S., Pomerleau, F., Liu, M., & Siegwart, R. (2013). 3D path planning and execution for search and rescue ground robots. In 2013 IEEE/RSJ international conference on intelligent robots and systems (IROS) (pp. 722–727). IEEE.

  • Diankov, R., Kuffner, J. (2007). Randomized statistical path planning. In IEEE/RSJ international conference on intelligent robots and systems. IROS 2007 (pp. 1–6). IEEE.

  • Douillard, B., Underwood, J., Kuntz, N., Vlaskine, V., Quadros, A., Morton, P., et al. (2011). On the segmentation of 3D lidar point clouds. In ICRA.

  • Du, T. C., Li, E. Y., & Chang, A. P. (2003). Mobile agents in distributed network management. Communications of the ACM, 46(7), 127–132.

    Article  Google Scholar 

  • Dubé, R., Dugas, D., Stumm, E., Nieto, J., Siegwart, R., & Cadena, C. (2017a). Segmatch: Segment based place recognition in 3D point clouds. In ICRA (pp. 5266–5272). IEEE.

  • Dubé, R., Gawel, A., Sommer, H., Nieto, J., Siegwart, R., & Cadena, C. (2017b). An online multi-robot slam system for 3D lidars. In IROS.

  • Elmaliach, Y., Agmon, N., & Kaminka, G. A. (2007). Multi-robot area patrol under frequency constraints. In ICRA (pp. 385–390).

  • Elmaliach, Y., Agmon, N., & Kaminka, G. A. (2009a). Multi-robot area patrol under frequency constraints. Annals of Mathematics and Artificial Intelligence, 57(3), 293–320.

    Article  MathSciNet  MATH  Google Scholar 

  • Elmaliach, Y., Agmon, N., & Kaminka, G. A. (2009b). Multi-robot area patrol under frequency constraints. Annals of Mathematics and Artificial Intelligence, 57(3–4), 293–320.

    Article  MathSciNet  MATH  Google Scholar 

  • Farinelli, A., Iocchi, L., & Nardi, D. (2004). Multirobot systems: A classification focused on coordination. IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics), 34(5), 2015–2028.

    Article  Google Scholar 

  • Farinelli, A., Iocchi, L., & Nardi, D. (2017). Distributed on-line dynamic task assignment for multi-robot patrolling. Autonomous Robots, 41(6), 1321–1345.

    Article  Google Scholar 

  • Ferri, F., Gianni, M., Menna, M., & Pirri, F. (2014). Point cloud segmentation and 3D path planning for tracked vehicles in cluttered and dynamic environments. In Proceedings of the 3rd IROS Workshop on Robots in Clutter: Perception and Interaction in Clutter.

  • Franchi, A., Freda, L., Oriolo, G., & Vendittelli, M. (2009). The sensor-based random graph method for cooperative robot exploration. IEEE/ASME Transaction on Mechatronics, 14(2), 163–175.

    Article  Google Scholar 

  • Grisetti, G., Kümmerle, R., Stachniss, C., & Burgard, W. (2010). A tutorial on graph-based slam. Intelligent Transportation Systems Magazine, IEEE, 2(4), 31–43.

    Article  Google Scholar 

  • Haït, A., Simeon, T., & Taïx, M. (2002). Algorithms for rough terrain trajectory planning. Advanced Robotics, 16(8), 673–699.

    Article  Google Scholar 

  • Hernández, E., Barrientos, A., & Cerro, J. D. (2014). Selective smooth fictitious play: An approach based on game theory for patrolling infrastructures with a multi-robot system. Expert Systems With Applications, 41(6), 2897–2913.

    Article  Google Scholar 

  • Hornung, A., Wurm, K. M., Bennewitz, M., Stachniss, C., & Burgard, W. (2013). OctoMap: An efficient probabilistic 3D mapping framework based on octrees. Autonomous Robots, 34(3), 189–206.

    Article  Google Scholar 

  • Iocchi, L., Marchetti, L., & Nardi, D. (2011). Multi-robot patrolling with coordinated behaviours in realistic environments. In IROS (pp. 2796–2801).

  • Jung, M. F., Beane, M., Forlizzi, J., Murphy, R., & Vertesi, J. (2017). Robots in group context: Rethinking design, development and deployment. In Proceedings of the 2017 CHI conference extended abstracts on human factors in computing systems (pp. 1283–1288). ACM.

  • Karaman, S., & Frazzoli, E. (2010). Incremental sampling-based algorithms for optimal motion planning. Robotics Science and Systems VI, 104, 2.

  • Kleiner, A., Heintz, F., & Tadokoro, S. (2016). Special issue on safety, security, and rescue robotics (SSRR), part 2. Journal of Field Robotics, 33(4), 409–410.

    Article  Google Scholar 

  • Kruijff, G. J. M., Kruijff-Korbayová, I., Keshavdas, S., Larochelle, B., Janíček, M., Colas, F., et al. (2014). Designing, developing, and deploying systems to support human-robot teams in disaster response. Advanced Robotics, 28(23), 1547–1570.

    Article  Google Scholar 

  • Kruijff, G. J. M., Pirri, F., Gianni, M., Papadakis, P., Pizzoli, M., Sinha, A., et al. (2012). Rescue robots at earthquake-hit Mirandola, Italy: A field report. In 2012 IEEE international symposium on safety, security, and rescue robotics (SSRR) (pp. 1–8). IEEE.

  • Kruijff-Korbayová, I., Colas, F., Gianni, M., Pirri, F., Greeff, J., Hindriks, K., et al. (2015). Tradr project: Long-term human-robot teaming for robot assisted disaster response. KI-Künstliche Intelligenz, 29(2), 193–201.

    Article  Google Scholar 

  • Kruijff-Korbayová, I., Freda, L., Gianni, M., Ntouskos, V., Hlaváč, V., Kubelka, V., et al. (2016). Deployment of ground and aerial robots in earthquake-struck amatrice in Italy (brief report). In 2016 IEEE international symposium on safety, security, and rescue robotics (SSRR) (pp. 278–279). IEEE.

  • Krüsi, P., Furgale, P., Bosse, M., & Siegwart, R. (2017). Driving on point clouds: Motion planning, trajectory optimization, and terrain assessment in generic nonplanar environments. Journal of Field Robotics, 34(5), 940–984.

    Article  Google Scholar 

  • Kubelka, V., Oswald, L., Pomerleau, F., Colas, F., Svoboda, T., & Reinstein, M. (2015). Robust data fusion of multimodal sensory information for mobile robots. Journal of Field Robotics, 32(4), 447–473.

    Article  Google Scholar 

  • LaValle, S. M. (2006). Planning algorithms. Cambridge: Cambridge University Press, Accessed Dec 2018.

  • Machado, A., Ramalho, G., Zucker, J. D., & Drogoul, A. (2002). Multi-agent patrolling: An empirical analysis of alternative architectures. In International workshop on multi-agent systems and agent-based simulation (pp. 155–170). Springer.

  • Menna, M., Gianni, M., Ferri, F., & Pirri, F. (2014). Real-time autonomous 3D navigation for tracked vehicles in rescue environments. In IROS (pp. 696–702).

  • Murphy, R. R. (2004). Trial by fire [rescue robots]. IEEE Robotics & Automation Magazine, 11(3), 50–61.

    Article  MathSciNet  Google Scholar 

  • Nagatani, K., Kiribayashi, S., Okada, Y., Otake, K., Yoshida, K., Tadokoro, S., et al. (2013). Emergency response to the nuclear accident at the fukushima daiichi nuclear power plants using mobile rescue robots. Journal of Field Robotics, 30(1), 44–63.

    Article  Google Scholar 

  • Panagou, D., Stipanovi, D. M., & Voulgaris, P. G. (2016). Distributed coordination control for multi-robot networks using lyapunov-like barrier functions. IEEE Transactions on Automatic Control, 61(3), 617–632.

    Article  MathSciNet  MATH  Google Scholar 

  • Park, C. H., Kim, Y. D., & Jeong, B. (2012). Heuristics for determining a patrol path of an unmanned combat vehicle. Computers & Industrial Engineering, 63(1), 150–160.

    Article  Google Scholar 

  • Pasqualetti, F., Durham, J. W., & Bullo, F. (2012). Cooperative patrolling via weighted tours: Performance analysis and distributed algorithms. IEEE Transactions on Robotics, 28(5), 1181–1188.

    Article  Google Scholar 

  • Pippin, C., & Christensen, H. (2014). Trust modeling in multi-robot patrolling. In ICRA (pp. 59–66).

  • Portugal, D. (2017). patrolling\(\_\)sim - Multi-Robot Patrolling Stage/ROS Simulation Package, Accessed February 20, 2017.

  • Portugal, D., & Rocha, R. (2010). Msp algorithm: Multi-robot patrolling based on territory allocation using balanced graph partitioning. In Proceedings of the 2010 ACM symposium on applied computing (pp. 1271–1276). New York, NY, USA: ACM.

  • Portugal, D., & Rocha, R. (2011). A survey on multi-robot patrolling algorithms. In Technological Innovation for Sustainability (pp. 139–146).

  • Portugal, D., & Rocha, R. P. (2013a). Distributed multi-robot patrol: A scalable and fault-tolerant framework. Robotics and Autonomous Systems, 61(12), 1572–1587.

    Article  Google Scholar 

  • Portugal, D., & Rocha, R. P. (2013b). Multi-robot patrolling algorithms: Examining performance and scalability. Advanced Robotics, 27(5), 325–336.

    Article  Google Scholar 

  • Portugal, D., & Rocha, R. P. (2013c). Retrieving topological information for mobile robots provided with grid maps (pp. 204–217). Berlin: Springer.

    Google Scholar 

  • Portugal, D., & Rocha, R. P. (2013d). Scalable, fault-tolerant and distributed multi-robot patrol in real world environments. In IROS (pp. 4759–4764).

  • Portugal, D., & Rocha, R. P. (2016). Cooperative multi-robot patrol with bayesian learning. Autonomous Robots, 40(5), 929–953.

    Article  Google Scholar 

  • Robin, C., & Lacroix, S. (2016). Multi-robot target detection and tracking: Taxonomy and survey. Autonomous Robots, 40(4), 729–760.

    Article  Google Scholar 

  • Rohmer, E., Singh, S. P. N., & Freese, M. (2013). V-rep: A versatile and scalable robot simulation framework. In Proceedings of The International Conference on Intelligent Robots and Systems (IROS).

  • Sak, T., Wainer, J., & Goldenstein, S. K. (2008). Probabilistic multiagent patrolling (pp. 124–133). Berlin: Springer.

  • Santana, H., Ramalho, G., Corruble, V., & Ratitch, B. (2004). Multi-agent patrolling with reinforcement learning. In Proceedings of the 3rd international joint conference on autonomous agents and multiagent systems—Volume 3, AAMAS’04 (pp. 1122–1129). IEEE Computer Society.

  • Schwarz, M. (2017). nimbro\(\_\)network - ROS transport for high-latency, low-quality networks, Accessed February 20, 2017.

  • Sempé, F., & Drogoul, A. (2003). Adaptive patrol for a group of robots. In 2003 IEEE/RSJ international conference on intelligent robots and systems. (IROS 2003). Proceedings (Vol. 3, pp. 2865–2869). IEEE.

  • Shahriari, M., & Biglarbegian, M. (2016). A new conflict resolution method for multiple mobile robots in cluttered environments with motion-liveness. IEEE Transactions on Cybernetics, PP(99), 1–12.

    Google Scholar 

  • Song, C., Liu, L., Feng, G., & Xu, S. (2014). Optimal control for multi-agent persistent monitoring. Automatica, 50(6), 1663–1668.

    Article  MathSciNet  MATH  Google Scholar 

  • Tardioli, D., Sicignano, D., Riazuelo, L., Romeo, A., Villarroel, J. L., & Montano, L. (2016). Robot teams for intervention in confined and structured environments. Journal of Field Robotics, 33(6), 765–801.

    Article  Google Scholar 

  • Walcott-Bryant, A., Kaess, M., Johannsson, H., & Leonard, J. J. (2012). Dynamic pose graph slam: Long-term mapping in low dynamic environments. In 2012 IEEE/RSJ international conference on intelligent robots and systems (IROS) (pp. 1871–1878). IEEE.

  • Weinmann, M., Jutzi, B., & Mallet, C. (2014). Semantic 3d scene interpretation: A framework combining optimal neighborhood size selection with relevant features. ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 2(3), 181.

    Article  Google Scholar 

  • Worst, R., Dubé, R., Svoboda, T., Freda, L., et al. (2017). Dr 6.3: Multi-robot task adaptation, TRADR deliverable. Accessed April 15, 2018.

  • Worst, R., Zimmermann, E., Reuter, D., et al. (2018). Dr 6.4: Persistence in long-term human-robot teaming for robot assisted disaster response, TRADR deliverable. Accessed October 13, 2018.

  • Yan, C., & Zhang, T. (2016). Multi-robot patrol: A distributed algorithm based on expected idleness. International Journal of Advanced Robotic Systems, 13(6), 1729881416663,666.

    Google Scholar 

  • Yan, Z., Jouandeau, N., & Cherif, A. A. (2013). A survey and analysis of multi-robot coordination. International Journal of Advanced Robotic Systems, 10(12), 399.

    Article  Google Scholar 

  • Yehoshua, R., Agmon, N., & Kaminka, G. A. (2013). Robotic adversarial coverage: Introduction and preliminary results. In IROS (pp. 6000–6005).

  • Zimmermann, K., Zuzanek, P., Reinstein, M., & Hlavac, V. (2014). Adaptive traversability of unknown complex terrain with obstacles for mobile robots. In 2014 IEEE international conference on robotics and automation (ICRA) (pp. 5177–5182). IEEE.

Download references


This work was supported by the European Union’s Seventh Framework Programme for research, technological development and demonstration under the TRADR Project No. FP7-ICT-609763.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Luigi Freda.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.



1.1 Code implementation

For the implementation of the patrolling agent algorithm, we used the C\({++}\) ROS package patrolling_sim as a starting point (Portugal 2017; Portugal and Rocha 2016). This is specifically designed for 2D patrolling tasks. It was used as a starting skeleton architecture providing core functionalities (such as graph management utilities). We significantly modified the core of this package in order to manage 3D data, implement our new patrolling agent algorithm, interface the agent module more tightly with the path planner and the 3D GUI in our network architecture.

An open source implementation of our framework is available.Footnote 27

1.2 Software design

A functional diagram of the presented multi-robot system is reported in Fig. 14. The main blocks are listed below.

The robots, each one with its own ID \(\in \{1,\ldots ,m\}\), have the same internal architecture and host the on-board functionalities which concern decision and processing aspects both at topological level and at metric level. According to Sect. 4.5, each robot maintains and updates an instance of the patrolling graph and of the metric map in its internal memory.

The core services, hosted in the main central computer, manage the multi-robot system persistence database and allow specific modules to load/save map, trajectories and patrolling graphs from/into the central database (for re-using relevant data along different missions).

The core modules, also hosted in central computer, include the patrolling graph builder and the patrolling monitor. The first builds a patrolling graph from a user assigned set of waypoints or from a saved history of robot trajectories. The built patrolling graph is then distributed to all the robots and saved in the central persistence database. The patrolling monitor continuously checks the current status of the patrolling activities and records relevant data for monitoring and benchmarking.

The multi-robot 3D GUI, hosted on one OCU (Operator Control Unit), is based on RVIZ and allows the user (i) to select multiple waypoints which can be fed to the path planners or to the patrolling graph builder (ii) to visualize relevant point cloud data, maps, and robot models (iii) to stop/restart robots when needed (iv) to trigger the loading/saving of maps and robot trajectories (v) to realign the current map of a selected robot to a loaded map.

The architecture is fully distributed without centralized coordination mechanisms. In particular, each robot hosts an instance of the patrolling agent and of the path-planner.

As shown in Fig. 14, the various modules in the architecture exchange different kind of messages. These are grouped in the following types.

  • Coordination messages These are exchanged amongst robots for sharing knowledge and decisions, in order to attain cooperation and coordination. For convenience, the patrol monitor records an history of these messages.

  • GUI messages These are exchanged with the 3D GUI and include both control messages and visualization data.

  • Load/save messages: these are exchanged with the core services and contain both loaded and saved data.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Freda, L., Gianni, M., Pirri, F. et al. 3D multi-robot patrolling with a two-level coordination strategy. Auton Robot 43, 1747–1779 (2019).

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: