Home /Research /An interacting multi-robot system and smart environment for studying collective behaviours
SWARM

An interacting multi-robot system and smart environment for studying collective behaviours

Rosie Ch Dawkins, Owen Holland, Alan Winfield, Phil Greenway, Arthur S. Stephens

Year
2002
Citations
13

Abstract

Complex collective behaviour, such as flock forming in birds, can be produced from the local interaction of simple agents. We have produced an environment for studying these emergent behaviours with multiple, interacting mobile robots. Each small, inexpensive, robot contains an embedded PC and a PCMCIA radio LAN. This provides high bandwidth communications and runs a wide range of PC software. Each robot estimates its position and orientation using odometry. Drift in this pose estimate is corrected by fusing with non-drifting data from video cameras and a tracking system. The pose is then broadcast to all other robots.

Keywords

OdometryRobotMobile robotComputer scienceComputer visionArtificial intelligenceBluetoothOrientation (vector space)Bandwidth (computing)Real-time computing

Related papers

Browse all SWARM papers