Home /Research /Faulty robot rescue by multi-robot cooperation
SWARM

Faulty robot rescue by multi-robot cooperation

Gyuho Eoh, Jeong Sik Choi, Beom H. Lee

Year
2013
Citations
12

Abstract

SUMMARY This paper presents a multi-agent behavior to cooperatively rescue a faulty robot using a sound signal. In a robot team, the faulty robot should be immediately recalled since it may seriously obstruct other robots, or collected matters in the faulty robot may be lost. For the rescue mission, we first developed a sound localization method, which estimates the sound source from a faulty robot by using multiple microphone sensors. Next, since a single robot cannot recall the faulty robot, the robots organized a heterogeneous rescue team by themselves with pusher, puller, and supervisor. This self-organized team succeeded in moving the faulty robot to a safe zone without help from any global positioning systems. Finally, our results demonstrate that a faulty robot among multi-agent robots can be immediately rescued with the cooperation of its neighboring robots and interactive communication between the faulty robot and the rescue robots. Experiments are presented to test the validity and practicality of the proposed approach.

Keywords

RobotRescue robotSearch and rescueComputer scienceArtificial intelligenceMicrophoneRobot controlSocial robotSimulationMobile robot

Related papers

Browse all SWARM papers