D. Fox, W. Burgard, H.Kruppa, and S. Thrun
Efficient Multi-RobotLocalization
Based on Monte Carlo Approximation
Proc. of the 9th International Symposium ofRobotics
Research (ISRR-99), 1999
Abstract
This paper presents a probabilistic algorithm for collaborative mobilerobot
localization. Our approach uses a sample-based version of Markovlocalization,
capable of localizing mobile robots in an any-time fashion.When teams of
robots localize themselves in the same environment, probabilisticmethods
are employed to synchronize each robot's belief whenever one robot detects
another. As a result, the robots localize themselves faster, maintainhigher
accuracy, and high-cost sensors are amortized across multiple robotplatforms.
The paper also describes experimental results obtained usingtwo mobile robots,
using computer vision and laser range-finding for detectingeach other and
estimating each other's relative location. The results, obtainedin an indoor
office environment, illustrate drastic improvements in localizationspeed
and accuracy when compared to conventional single-robot localization.
Download
Full paper
[.ps.gz](1268 kb, 8 pages)
Bibtex
@InCollection{Fox00Eff,
author = {Fox, D. and Burgard, W. and Kruppa,H. and
Thrun, S.},
title = {Efficient multi-robot localizationbased on
{M}onte {C}arlo approximation},
booktitle = {Robotics Research: the Ninth InternationalSymposium},
publisher = {Springer-Verlag},
year = 2000,
editor = {Hollerbach, J. and Koditschek, D.},
address = {London}
}