This Article Full Text (PDF) References Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in Web of Science Alert me to new issues of the journal Add to Saved Citations Download to citation manager Request Permissions Request Reprints Add to My Marked Citations Citing Articles Citing Articles via Google Scholar Citing Articles via Scopus Google Scholar Articles by Starke, J. Articles by Fukuda, T. Search for Related Content Social Bookmarking                         What's this?

Experimental Study on Self-organized and Error Resistant Control of Distributed Autonomous Robotic Systems

Interdisciplinary Center for Scientific Computing and Institute of Applied Mathematics, University of Heidelberg, D-69120 Heidelberg, Germany, starke{at}iwr.uni-heidelberg.de

T. Kaga

Department of Micro System Engineering, Nagoya University, Nagoya 464-8603, Japan, kaga{at}robo.mein.nagoya-u.ac.jp

M. Schanz

Institute of Parallel and Distributed Systems (IPVS), D-70569 Stuttgart, Germany, michael.schanz{at}informatik.uni-stuttgart.de

T. Fukuda

Center of Cooperative Research in Advanced Science and Technology, Nagoya University, Nagoya 464-8603, Japan, fukuda{at}mein.nagoya-u.ac.jp

The assignment of distributed mobile autonomous robots to targets, which occurs for instance as an important task in flexible manufacturing environments, is solved by using a self-organization approach motivated by pattern formation principles in biological, chemical, and physical systems. Similar to observations in many natural systems, such as ant tribes, the pattern formation of colored shells or convection patterns in the Rayleigh-Bénard problem of fluid dynamics, the self-organization principles lead to a robust and fault tolerant behavior where the patterns or structures recover from disturbances. The considered problem is the dynamic assignment of a number of robots to given targets where the mobile robots have to move to the targets in order to perform some tasks there. Hereby, each robot uses only local information (i.e., no world coordinate system is necessary). The underlying mathematical problem of the robot-target assignment is the so-called two-index assignment problem from combinatorial optimization. The approach used guarantees always feasible solutions in the assignment of robotic units to targets. As a consequence, for scenarios with only convex obstacles with large enough distances to each other, no spurious states cause the assignment process to fail. The error resistant control method for distributed autonomous robotic systems is demonstrated by several experiments with mobile robots. These results are compared and supplemented with computer simulations.

Key Words: self-organization • selection equation • distributed robotic systems • assignment problems

The International Journal of Robotics Research, Vol. 24, No. 6, 465-486 (2005)
DOI: 10.1177/0278364905053239


CiteULike    Complore    Connotea    Del.icio.us    Digg    Reddit    Technorati    Twitter    What's this?