ECAI 2004 Conference Paper

[full paper]

Adaptive Robot Coordination using Interference Metrics

Avi Rosenfeld, Sarit Kraus, Gal Kaminka

One main issue facing robotic groups is effective coordination mechanisms. Many robotic domains contain restrictions such as limited areas of operation or narrow passages that are liable to cause spacial conflicts between robots and their environment. Previous work presented a novel definition of interference that measured the total time robots deal with resolving such conflicts over space. It was found that a robotic group's productivity was strongly negatively correlation to this measure. Effective coordination techniques minimize this level of interference and thus achieve higher productivity. This paper uses this result towards creating adaptive coordination techniques that are able to dynamically adjust to the level of interference a given robot in the group faces. Using the robot's internal interference metric as a guide, we are able to create coordination methods that can quickly and effectively adjust to a given domain's spacial limitations. We present a heuristic that is completely distributed and requires no communication between robots. We found that groups that used this approach consistently achieved higher productivity than non-adaptive coordination methods.

Keywords: Cognitive Robotics, Robotics

Citation: Avi Rosenfeld, Sarit Kraus, Gal Kaminka: Adaptive Robot Coordination using Interference Metrics. In R.López de Mántaras and L.Saitta (eds.): ECAI2004, Proceedings of the 16th European Conference on Artificial Intelligence, IOS Press, Amsterdam, 2004, pp.910-914.