` Intelligent and Mobile Robotics

Swarm Robotics

The research of swarms at IMR aims to integrate principles and theoretical background of swarm behaviours with methodology/theory describing cooperative localization of autonomous robots and principles of self-organizing adaptation leading to a flexible stand-alone system. It will enable applicability of SWARM robotics in realistic outdoor scenarios of surveillance and reconnaissance. Basically, we develop principles of a decentralized relative localization of neighboring particles that are integrated to swarm behaviors with an aim to keep reciprocal visibility between neighbors. This enables to employ micro-UAV swarms outside laboratories equipped by a precise positioning system. Besides, a concept of adaptively evolving swarm behaviors is established to decrease relative localization uncertainty. To enable multi-robot applications, theoretical principles of determining desired shapes of micro-UAV swarms are designed based on bio-inspired methods of artificial intelligence, namely Self-Organizing Maps and Particle Swarm Optimization. Finally, decentralized collective decision making mechanisms are established with a theory identifying necessary assumptions of the switching between different swarm behaviors. This research is aimed at a study of observed autonomous behaviors of micro-UAV swarms.

As a parallel stream, we are developing principles of autonomous formation driving and stabilization into a desired shape of the group. Our research is focused on mobile robot coordination in environment with dynamic obstacles, on splitting and coupling of teams and on trajectory planning for the formations. From the application perspective, the research is aimed at road sweeping, particularly at airport snow shoveling and park pathways maintenance.

Finally, we investigate methods of coordination of heterogeneous teams (UAVs&UGVs) including autonomous take-off and landing. This research is focused on surveillance, where UAV is employed to survey areas inaccessible for UGV.

People involved:

Libor Přeučil, Jan Chudoba

Contact:

Libor Přeučil

Demos:

3D simulation of swarm movement using the escape behavior method

This movie presents an investigation of swarm control dealing with an escape behavior, which is important functionality in application with human-swarm coexistence. The escape behavior algorithm was extended for the swarm purposes. The movement strategies originally developed for holonomic point particles were replaced with dynamic models of UAVs. Examples of the swarm movement under the rules of the escape behavior using the dynamic models of UAVs are shown in the movie.




Navigation, localization and stabilization of formations of unmanned aerial and ground vehicles

A leader-follower formation driving algorithm is employed for control of heterogeneous groups of unmanned micro aerial and ground vehicles stabilized under a top-view relative localization in this video. The core of the method lies in an avoidance function, in which the entire 3D formation is represented by a convex hull projected along a desired path to be followed by the group. Such a representation of the formation provides non-collision trajectories of the robots and respects requirements of the direct visibility between the team members in environment with static as well as dynamic obstacles, which is crucial for the top-view localization. The algorithm is suited for utilization of a simple yet stable visual based navigation of the group (referred to as GeNav), which together with the on-board relative localization enables deployment of large teams of micro-scale robots in environments without any available global localization system.




Ar.Drone in robotic research, autonomous flight, autonomous navigation and formation driving

Several navigation tasks utilizing a low-cost Micro Aerial Vehicle (MAV) platform AR-drone are presented in this video to show how it can be used in an experimental verification of scientific theories and developed methodologies. The presented methods rely on visual navigation and localization using on-board cameras of the ARdrone employed in the control feedback. The aim of this video is to demonstrate flight performance of this platform in real world scenarios of mobile robotics.




Cooperative UAV-UGV inspection

The video demonstrates a heterogenous UAV-UGV system in an autonomous inspection task. The mission is to visit a set of predefined places. The usual problem of inspection tasks is that while ground robots cannot access all areas, small UAVs are limited by their low flying time and payload. The UAV-UGV team is able to overcome these limitations. The UAV is launched whenever the inspected area seems to be inaccessible. The ground robot heliport is able to adjust the quadrotor position after landing, which will allow to recharge the UAV.




Experiment of swarm movement using the escape behavior method with ground robots

The movie shows principles of swarm control based on the escape behavior method verified by experiments with ground robots. The ground robots are used for verification of developed theoretical principles and methods, before the finalization of the development of unmanned aerial vehicles control. The robots know their positions and positions of dynamic obstacles in the environment.




3D simulation of swarm coverage using the PSO algorithm for the shape optimization.

Particle Swarm Optimization (PSO) was employed for the swarm control with the aim to optimize locations of UAVs in a surveillance mission. The swarm particles are forced to observe areas of interest with different priority. During the swarm movement restrictions of direct visibility in limited radius is kept (the lines connecting the UAVs in the movie).




Airport snow shoveling

The video shows several experiments with shoveling snow by a formation of mobile robots. The path for the formation (and all its robots) is planned using Receding Horizon Control approach. In the first part of the video several simulations show ability to cope with coupling/decapling of the formation and dynamic obstacle avoidance. Then two indoor experiments show how to turn the formation in a blind road (e.g. at the end of a runway) and how to shovel a snow in an airport. The last experiment with P3AT robot shows P3AT mobile robot shoveling a pavement in a park. In the last experiment the robot is navigated by SURFnav algorithm.




3D simulation of swarm movement according to PSO rules.

Evolving swarm behavior established using evolutionary principles to solve applications of cooperative searching in the 3D environment. Particle Swarm Optimization (PSO) algorithm was extended for the swarm purposes. The movement strategies originally developed for holonomic point particles were replaced with dynamic models of UAVs. In comparison with the classical PSO algorithm, an obstacle avoidance mechanism was included with focus on relative interactions between the particles caused by the air flow from rotors of real helicopters. Besides, the influence of the model of relative localization introduced into the algorithm through stabilization constraints was studied. It was shown that the stabilization constraints are sufficient to ensure the reciprocal visibility between neighboring particles in PSO. It was also observed that the stabilization constraints may positively influence the performance of the PSO. In movie, an origin of several sub-swarms may be observed during PSO searching for the place of the highest intensity of an electromagnetic field surrounding transmitters. Each of the incidentally created groups searches around a promising local extreme of the intensity. As the evolution of the PSO continues, the less successful swarms are autonomously merged with the groups searching around an extreme with higher intensity. This behavior was not pre-programmed, but it appeared emergently.




Experiment of swarm movement according to PSO rules with ground robots.

Principles of swarm control based on the PSO verified by experiments with ground robots. The ground robots are used for verification of developed theoretical principles and methods, before the finalization of the development of unmanned aerial vehicles control. The robots know only the intensity of the gray color in the locations, where they are. The aim is to find the whitest place.




Coordination and navigation of heterogeneous UAVs-UGVs teams localized by a hawk-eye approach

A navigation and stabilization approach for 3D heterogeneous (UAVs and UGVs) formations acting under a hawk-eye like relative localization is employed in this paper. A Model Predictive Control (MPC) based concept is used for the formation driving in a leader-follower constellation into a required target region. The formation to target region problem in 3D is solved using the MPC methodology for both: i) the trajectory planning and control of a virtual leader, and ii) the control and stabilization of followers - UAVs and UGVs.


See IMR demos page for more records of experiments of the SWARM stream and IMR in general.

See AR-drone robotic research page for movies and description of research with AR-drone quadrotor platform at Intelligent and Mobile Robotics Group of Czech Technical University in Prague.

Experimental platforms and facilities:

MikroKopter L4-ME:

The main experimental platform employed for verification of the swarm experiments is the L4-ME quadrotor designed by the MikroKopter company. We are building a fleet of >5 UAVs equipped with relative visual localization.


Module of visual relative localization

This device is the key component of the arising system of swarm control based on the visual relative localization feedback. The small, light-weight, low-cost, fast and reliable system is based on off the shelf components consisting of the Caspa camera module and Gumstix Overo board accompanied by a developed efficient image processing method for detecting black and white circular patterns. The developed system exhibits reliable and fast estimation of the relative position of the pattern up to 30 fps using the full resolution of the Caspa camera. Details on the system can be found in: ICRA 2013 paper, [bib].


AR-drone

A lighter quadrotor of the Parrot company is employed for experiments in environments with possible occurrence of humans and for testing algorithms of human-UAV interaction. In our system, the AR-drone is capable of landing on a UGV to form a heterogenous team. Details on the AR-drone control system for autonomous flying, which is available for free by our team, may be found in: IROS 2012 paper, [bib].


Pioneer 3-AT with a mobile helipad

We have developed a cognitive mobile helipad being able to change its shape and support such the landing UAV and UAV carrying by the robot. Details on the landing and take-off maneuvers and the helipad design may be found in IEEE SSD 2012 paper, [bib].


Outdoor testing arena

To fulfill regulations given by the airspace authority and to protect UAVs in swarm experiments, we have built a large net, which enables outdoor experiments.


Cable driven parallel multiple manipulator to emulate a swarm of UAVs

This platform was developed for a preliminary testing of the relative visual navigation between swarm entities. The main idea of the robot is to enable precise positioning of multiple objects in 3D. This allows us to safely demonstrate swarming behaviors gather meaningful data for the development of visual relative localization. Details on the robot design and basic planning algorithms are provided in: Vonásek, V. - Saska, M. - Přeučil, L. Motion Planning for a Cable Driven Parallel Multiple Manipulator Emulating a Swarm of MAVs. Proceedings of the 9th International Workshop on Robot Motion Control, RoMoCo'13, 2013.


Multi-robot SyRoTek platform

The SyRoTek (“System for robotic e-learning”) allows to remotely (via internet) control up to 13 robots in a dynamic environment and monitor their behaviour on-line during real experiments. Since the system is 24/7 ready to use, it is an ideal platform for testing swarm and formation driving algorithms. Details on the SyRoTek system can be found at SyRoTek web page.


Selected publications:

Journal articles, books, book chapters:

  • Saska, M. - Vonásek, V. - Přeučil, L.: Trajectory Planning and Control for Airport Snow Sweeping by Autonomous Formations of Ploughs, Journal of Intelligent & Robotic Systems, April, 2013. DOI 10.1007/s10846-013-9829-3. (In Press) [pdf]
  • Saska, M. - Mejía, J. S. - Stipanović, D. M. - Vonásek, V. - Schilling, K. - Přeučil, L.: Control and navigation in manoeuvres of formations of unmanned mobile vehicles, European Journal of Control, 2013. (In Press)
  • Saska, M.: Trajectory planning and optimal control for formations of autonomous robots. Schriftenreihe Würzburger Forschungsberichte in Robotik und Telematik, Band 3. Würzburg: Universität Würzburg. 2010. ISSN: 1868-7466
  • Hess, M. - Saska, M. - Schilling, K.: Application of Coordinated Multi Vehicle Formations for Snow Shoveling on Airports. Inteligent Service Robotics, Volume 2, Number 4, Pages 205-217, October, 2009

Conference papers:

  • Saska, M. - Krajník, T. - Vonásek, V. - Přeučil, L.: Navigation, localization and stabilization of formations of unmanned aerial and ground vehicles. In Proceedings of the International Conference on Unmanned Aircraft Systems (ICUAS'13), 2013.
  • Faigl, J. - Krajník, T. - Chudoba, J. - Přeučil, L. - Saska, M.: Low-Cost Embedded System for Relative Localization in Robotic Swarms, Proc. of IEEE International Conference on Robotics and Automation (ICRA), 2013. [pdf] [bib]
  • Saska, M. - Vonásek, V. - Krajník, T. - Přeučil, L.: Coordination and Navigation of Heterogeneous UAVs-UGVs Teams Localized by a Hawk-Eye Approach. In Proceedings of 2012 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2012), 2012. [pdf] [bib]
  • Saska, M. - Krajník, T. - Faigl, J. - Vonásek, V. - Přeučil, L.: Low Cost MAV Platform AR-Drone in Experimental Verifications of Methods for Vision Based Autonomous Navigation. In Proceedings of IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2012), 2012. [pdf] [bib]
  • Saska, M. - Krajník, T. - Přeučil, L.: Cooperative Micro UAV-UGV Autonomous Indoor Surveillance. In International Multi-Conference on Systems, Signals and Devices (IEEE SSD 2012), 2012. [pdf] [bib]
  • Saska, M. - Vonásek, V. - Přeučil, L.: Roads Sweeping by Unmanned Multi-vehicle Formations. In IEEE International Conference on Robotics and Automation (ICRA 2011), 2011. [pdf] [bib]
  • M. Saska, V. Vonasek, T. Krajnik. Airport snow shoveling. In IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2010), 2010. [pdf] [bib]
  • M. Saska, V. Vonasek, L. Preucil. Control of ad-hoc formations for autonomous airport snow shoveling. In IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2010), 2010. [pdf] [bib]
  • M. Saska, M. Hess and K. Schilling. Efficient Airport Snow Shoveling by Applying Autonomous Multi-Vehicle Formations. IEEE International Conference on Robotics and Automation (ICRA 2008), 2008. [pdf] [bib]
  • M. Saska, J. S. Mejia, D. M. Stipanovic, and K. Schilling. Control and navigation of formations of car-like robots on a receding horizon, in IEEE Control Applications, (CCA) & Intelligent Control, (ISIC), part of the IEEE Multi-Conference on Systems & Control (IEEE MSC), 2009. [pdf] [bib]
  • Saska, M. - Vonásek, V. - Přeučil, L.: Formation Coordination with Path Planning in Space of Multinomials. In Artificial Intelligence and Soft Computing, 2011. [pdf] [bib]
  • M. Saska, V. Vonasek, L. Preucil. Navigation and Formation Control Employing Complementary Virtual Leaders for Complex Maneuvers. In 7th international Conference on Informatics in Control, Automation and Robotics, 2010. [pdf] [bib]
  • M. Saska, I. Ferenczi, M. Hess and K. Schilling. PATH PLANNING FOR FORMATIONS USING GLOBAL OPTIMIZATION WITH SPARSE GRIDS. In Proceedings of The 13th IASTED International Conference on Robotics and Applications (RA 2007), Wuerzburg, Germany, 2007. [pdf] [bib]

Student works

  • Zdeněk Kasl - 3D formations of unmanned aerial vehicles (M.Sc. thesis supervised by Martin Saska)
  • Tomáš Báča - Control of relatively localized unmanned helicopters (Bc. thesis supervised by Martin Saska)
  • Jan Langr - Odor source localization using swarm of unmanned helicopters (Bc. thesis supervised by Martin Saska)
  • Jindřich Mráček - FSS algorithm adapted for control of swarm of unmanned helicopters (Bc. thesis supervised by Martin Saska)
  • Vojtěch Spurný - Heterogeneous formations of ground vehicles and unmanned helicopters (Bc. thesis supervised by Martin Saska)
  • Adam Třešňák - Shape optimization of swarm of unmanned helicopters (Bc. thesis supervised by Martin Saska)
  • Jan Vakula - Escape behavior in swarms of unmanned helicopters [pdf] (Bc. thesis supervised by Martin Saska)
  • Filip Eckstein - Formation control in environment with dynamic obstacles [pdf] (Bc. thesis supervised by Martin Saska)
  • Vojtěch Pavlík - Swarm intelligence applied in multi-robot applications [pdf] (Bc. thesis supervised by Martin Saska)
  • Pavel Zedník - Relative visual localization in swarms of unmanned aerial vehicles [pdf] (Bc. thesis supervised by Martin Saska)

Projects and funding:

COLOS - grant MŠMT no. LH11053
Symbrion-Enlarged - European Union grant no. 216342
PostDoc GAČR - Martin Saska - grant MŠMT no. P103/12/P756