pyswarming

pyswarming is a research toolkit for Swarm Robotics.

Installation

You can install pyswarming from PyPI using pip (Recommended):

pip install pyswarming

Dependencies

pyswarming's dependencies are: numpy, numdifftools and matplotlib.

Documentation

The official documentation is hosted on ReadTheDocs.

Algorithms covered

This library includes the following algorithms to be used in swarm robotics:

• Leaderless heading consensus: the collective performs heading consensus ¹;
• Inverse power: ajustable attraction and repulsion laws ²;
• Spring: allows the robots to maintain a desired distance between them ²;
• Force law: mimics the gravitational force ³;
• Repulsive force: makes the individuals repulse each other ⁴;
• Body force: introduces a body force that considers the radii of the robots ⁴;
• Inter robot spacing: allows the robots to maintain a desired distance between them ⁵;
• Dissipative: a dissipative force that reduces the "energy" of the robots ⁵;
• Leader following: the collective performs heading consensus with a leader ⁶;
• Collision avoidance: the robot stays away from neighbors in the vicinity ⁷;
• Attraction alignment: the robot becomes attracted and aligned ⁷;
• Preferred direction: the robot has a preference to move toward a preset direction ⁷;
• Lennard-Jones: allows the formation of lattices ⁸;
• Virtual viscosity: a viscous force that reduces the "oscillation" of the robots ⁸;
• Modified attraction alignment: the robot becomes attracted and aligned by considering a “social importance” factor ⁹;
• Heading consensus: the collective performs heading consensus ¹⁰;
• Perimeter defense: the robots maximize the perimeter covered in an unknown environment ¹⁰;
• Environment exploration: provides spatial coverage ¹⁰;
• Aggregation: makes all the individuals aggregate collectively ¹¹;
• Alignment: the collective performs heading consensus ¹¹;
• Geofencing: attract the robots towards area A ¹¹;
• Repulsion: makes all the individuals repulse collectively ¹¹;
• Target: the robot goes to an specific target location ¹¹;
• Area coverage: using the Geofencing and Repulsion algorithms ¹¹;
• Collective navigation: using the Target and Repulsion algorithms ¹¹;
• Flocking: using the Aggregation, Repulsion and Alignment algorithms ¹¹;

Citing PySwarming

If you make use of PySwarming for your research, please cite our JOSS publication. Here is the corresponding BibTeX entry:

@article{deAndrade2023,
    doi = {10.21105/joss.05647},
    url = {https://doi.org/10.21105/joss.05647},
    year = {2023},
    publisher = {The Open Journal},
    volume = {8},
    number = {89},
    pages = {5647},
    author = {Emerson Martins de Andrade and Antonio Carlos Fernandes and Joel Sena Sales},
    title = {PySwarming: a research toolkit for Swarm Robotics},
    journal = {Journal of Open Source Software}
}

Examples using pyswarming.swarm

# importing the swarm creator
import pyswarming.swarm as ps

Repulsion

# creating the swarm
my_swarm = ps.Swarm(n = 10, # number of robots
                    linear_speed = 0.5, # linear speed of each robot
                    dT = 1.0, # sampling time
                    deployment_point_limits = [[0.0, 0.0, 0.0], [5.0, 5.0, 0.0]], # lower and upper limits for the position deployment
                    deployment_orientation_limits = [[0.0, 0.0, 0.0], [0.0, 0.0, 2*3.1415]], # lower and upper limits for the orientation deployment
                    distribution_type =  'uniform', # type of distribution used to deploy the robots
                    plot_limits = [[-50.0, 50.0], [-50.0, 50.0]], # plot limits x_lim, y_lim
                    behaviors = ['repulsion']) # list of behaviors
my_swarm.simulate()

Collective navigation

# creating the swarm
my_swarm = ps.Swarm(n = 10, # number of robots
                    linear_speed = 0.5, # linear speed of each robot
                    dT = 1.0, # sampling time
                    deployment_point_limits = [[0.0, 0.0, 0.0], [5.0, 5.0, 0.0]], # lower and upper limits for the position deployment
                    deployment_orientation_limits = [[0.0, 0.0, 0.0], [0.0, 0.0, 2*3.1415]], # lower and upper limits for the orientation deployment
                    distribution_type =  'uniform', # type of distribution used to deploy the robots
                    plot_limits = [[-50.0, 50.0], [-50.0, 50.0]], # plot limits x_lim, y_lim
                    behaviors = ['collective_navigation']) # list of behaviors
my_swarm.behaviors_dict['r_out']['collective_navigation']['alpha'] = 2.0  # setting the strength of the repulsion
my_swarm.behaviors_dict['r_out']['collective_navigation']['T'] = [-40, -40, 0] # setting the target
my_swarm.simulate()

Target + Aggregation

# creating the swarm
my_swarm = ps.Swarm(n = 10, # number of robots
                    linear_speed = 0.5, # linear speed of each robot
                    dT = 1.0, # sampling time
                    deployment_point_limits = [[0.0, 0.0, 0.0], [5.0, 5.0, 0.0]], # lower and upper limits for the position deployment
                    deployment_orientation_limits = [[0.0, 0.0, 0.0], [0.0, 0.0, 2*3.1415]], # lower and upper limits for the orientation deployment
                    distribution_type =  'uniform', # type of distribution used to deploy the robots
                    plot_limits = [[-50.0, 50.0], [-50.0, 50.0]], # plot limits x_lim, y_lim
                    behaviors = ['target','aggregation']) # list of behaviors
my_swarm.behaviors_dict['r_out']['target']['T'] = [-40, -40, 0] # setting the target
my_swarm.simulate()

Other Examples

Considering a swarm of robots, they can show different behaviors by using pyswarming. The following codes are simplified implementations, for detailed ones, see the examples folder.

# importing the swarming behaviors
import pyswarming.behaviors as pb

# importing numpy to work with arrays
import numpy as np

Target

To simplify, considering just one robot.

# define the robot (x, y, z) position
robot_position_i = np.asarray([0., 0., 0.])

# set the robot speed
robot_speed_i = 1.0

# define a target (x, y, z) position
target_position = np.asarray([8., 8., 0.])

for t in range(15):

    # print the robot (x, y, z) position
    print(robot_position_i)

    # update the robot (x, y, z) position
    robot_position_i += robot_speed_i*pb.target(robot_position_i, target_position)

Aggregation

Considering four robots.

# define each robot (x, y, z) position
robot_position = np.asarray([[8., 8., 0.],
                            [-8., 8., 0.],
                            [8., -8., 0.],
                            [-8., -8., 0.]])

# set the robot speed
robot_speed = 1.0

for time_i in range(15):

    # print the robot (x, y, z) positions
    print(robot_speed)

    # update the robot (x, y, z) positions
    for r_ind in range(len(robot_speed)):
        r_i = robot_speed[r_ind]
        r_j = np.delete(robot_speed, np.array([r_ind]), axis=0)
        robot_speed[r_ind] += robot_speed*pb.aggregation(r_i, r_j)

Repulsion

Considering four robots.

# define each robot (x, y, z) position
robot_position = np.asarray([[1., 1., 0.],
                            [-1., 1., 0.],
                            [1., -1., 0.],
                            [-1., -1., 0.]])

# set the robot speed
robot_speed = 1.0

for time_i in range(15):

    # print the robot (x, y, z) positions
    print(robot_position)

    # update the robot (x, y, z) positions
    for r_ind in range(len(robot_position)):
        r_i = robot_position[r_ind]
        r_j = np.delete(robot_position, np.array([r_ind]), axis=0)
        robot_position[r_ind] += robot_speed*pb.repulsion(r_i, r_j, 3.0)

Aggregation + Repulsion

Considering four robots.

# define each robot (x, y, z) position
robot_position = np.asarray([[8., 8., 0.],
                            [-8., 8., 0.],
                            [8., -8., 0.],
                            [-8., -8., 0.]])

# set the robot speed
robot_speed = 1.0

for time_i in range(15):

    # print the robot (x, y, z) positions
    print(robot_position)

    # update the robot (x, y, z) positions
    for r_ind in range(len(robot_position)):
        r_i = robot_position[r_ind]
        r_j = np.delete(robot_position, np.array([r_ind]), axis=0)
        robot_position[r_ind] += s_i*(pb.aggregation(r_i, r_j) + pb.repulsion(r_i, r_j, 5.0))

Contributing to pyswarming

All kind of contributions are welcome:

• Improvement of code with new features, bug fixes, and bug reports
• Improvement of documentation
• Additional tests

Follow the instructions here for submitting a PR (Pull Request).

If you have any ideas or questions, feel free to open an issue.

Star History

Acknowledgements

The authors would like to thank the Programa de Recursos Humanos da Agência Nacional do Petróleo, Gás Natural e Biocombustíveis (PRH18-ANP) for their financial support, supported with resources from investment by oil companies qualified in the P, DI Clause of ANP Resolution no. 50/2015. This work was supported by "Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - Brasil (CAPES)", LOC/COPPE/UFRJ (Laboratory of Waves and Current - Federal University of Rio de Janeiro) and the National Council for Scientific and Technological Development (CNPq), which are gratefully acknowledged.

Footnotes

1. T. Vicsek, A. Czirók, E. Ben-Jacob, I. Cohen, and O. Shochet, “Novel Type of Phase Transition in a System of Self-Driven Particles,” Phys. Rev. Lett., vol. 75, no. 6, pp. 1226–1229, Aug. 1995. https://doi.org/10.1103/PhysRevLett.75.1226. ↩

2. J. H. Reif and H. Wang, “Social potential fields: A distributed behavioral control for autonomous robots,” Robot. Auton. Syst., vol. 27, no. 3, pp. 171–194, May 1999. https://doi.org/10.1016/S0921-8890(99)00004-4. ↩ ↩²

3. W. M. Spears and D. F. Gordon, “Using artificial physics to control agents,” in Proceedings 1999 International Conference on Information Intelligence and Systems (Cat. No.PR00446), Bethesda, MD, USA: IEEE Comput. Soc, 1999, pp. 281–288. https://doi.org/10.1109/ICIIS.1999.810278. ↩

4. D. Helbing, I. Farkas, and T. Vicsek, “Simulating dynamical features of escape panic,” Nature, vol. 407, no. 6803, pp. 487–490, Sep. 2000. https://doi.org/10.1038/35035023. ↩ ↩²

5. N. E. Leonard and E. Fiorelli, “Virtual leaders, artificial potentials and coordinated control of groups,” presented at the IEEE Conference on Decision and Control, 2001. https://doi.org/10.1109/CDC.2001.980728. ↩ ↩²

6. A. Jadbabaie, Jie Lin, and A. S. Morse, “Coordination of groups of mobile autonomous agents using nearest neighbor rules,” IEEE Trans. Autom. Control, vol. 48, no. 6, pp. 988–1001, Jun. 2003. https://doi.org/10.1109/TAC.2003.812781. ↩

7. I. D. Couzin, J. Krause, N. R. Franks, and S. A. Levin, “Effective leadership and decision-making in animal groups on the move,” Nature, vol. 433, no. 7025, pp. 513–516, Feb. 2005. https://doi.org/10.1038/nature03236. ↩ ↩² ↩³

8. C. Pinciroli et al., “Lattice Formation in Space for a Swarm of Pico Satellites,” in Ant Colony Optimization and Swarm Intelligence, M. Dorigo, M. Birattari, C. Blum, M. Clerc, T. Stützle, and A. F. T. Winfield, Eds., in Lecture Notes in Computer Science, vol. 5217. Berlin, Heidelberg: Springer Berlin Heidelberg, 2008, pp. 347–354. https://doi.org/10.1007/978-3-540-87527-7_36. ↩ ↩²

9. R. Freeman and D. Biro, “Modelling Group Navigation: Dominance and Democracy in Homing Pigeons,” J. Navig., vol. 62, no. 1, pp. 33–40, Jan. 2009. https://doi.org/10.1017/S0373463308005080. ↩

10. M. Chamanbaz et al., “Swarm-Enabling Technology for Multi-Robot Systems,” Front. Robot. AI, vol. 4, Apr. 2017. https://doi.org/10.3389/frobt.2017.00012. ↩ ↩² ↩³

11. B. M. Zoss et al., “Distributed system of autonomous buoys for scalable deployment and monitoring of large waterbodies,” Auton. Robots, vol. 42, no. 8, pp. 1669–1689, Dec. 2018. https://doi.org/10.1007/s10514-018-9702-0. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸