Distributed 3D Encirclement with Target Localization and Collision Avoidance

Submitted by Antonio Franchi on

ImageThis video shows simulations and experiments of a 3D encirclement strategy for collective control of multiple robots (both 2D and 3D).

The ultimate version of the method can be found in [1].

The video below shows an implementation of the encirclement controller proposed in [2]. performed with five Khepera III robots; each robot is also equipped with a wireless card and a Hokuyo URG-04LX laser rangefinder. One of the robots acts as a the stationary target. The first video shows the live scene as well as the estimates computed by the four robots used for encirclement. In the first phase of the experiment, the robots reach a regular formation. Then one of the robot is kidnapped and released in a displaced position. The encirclement system is very quick to react and recover the correct formation. In the final phase of the experiment, two more robots in succession are kidnapped and powered off, and the remaining robots arrange themselves first as a 3-robot and then as a 2-robot regular formation. The second video shows another experiment in which three robots encircle a moving target. We obtained satisfactory results as long as the speed of the target remains at least one order of magnitude smaller than that of the robots.

References

  1. A. Franchi, Stegagno, P., and Oriolo, G., “Decentralized Multi-Robot Encirclement of a 3D Target with Guaranteed Collision Avoidance”, Autonomous Robots, vol. 40, no. 2, pp. 245-265, 2016.
  2. A. Franchi, Stegagno, P., Di Rocco, M., and Oriolo, G., “Distributed Target Localization and Encirclement with a Multi-robot System”, in 7th IFAC Symposium on Intelligent Autonomous Vehicles, Lecce, Italy, 2010.