We present a novel method for robot path following in an RFID-equipped corridor using received signal strength (RSS). RSS measurements are recorded as reference fingerprints during the exploration phase. These data are used to guide the robot through the corridor in the navigation phase. Distance and orientation that the robot derivates from the expected trajectory are approximated through the RSS differences and the index differences of two antennas. The RSS differences are calculated by comparing the current fingerprint to reference fingerprints, and the index differences are estimated by a particle filter. Finally, a P-controller is used for the navigation of the robot. This enables the robot to follow a certain path, which may be defined in the first phase of our approach. The actual experiments with a SCITOS G5 service robot confirm the validity and robustness of our approach.
[pdf]
@inproceedings{liu1109:Path,
author = {Ran Liu and Philipp Vorst and Artur Koch and Andreas Zell},
title = {Path Following for Indoor Robots with {RFID} Received Signal Strength},
booktitle = {The 19th International Conference on Software, Telecommunications
and Computer Networks (SoftCOM 2011)},
year = {2011},
address = {Split, Hvar, and Dubrovnik, Croatia},
month = {September},
note = {(Best paper award at the Symposium on RFID Technologies and Internet
of Things)},
abstract = {We present a novel method for robot path following in an RFID-equipped
corridor using received signal strength (RSS). RSS measurements are
recorded as reference fingerprints during the exploration phase.
These data are used to guide the robot through the corridor in the
navigation phase. Distance and orientation that the robot derivates
from the expected trajectory are approximated through the RSS differences
and the index differences of two antennas. The RSS differences are
calculated by comparing the current fingerprint to reference fingerprints,
and the index differences are estimated by a particle filter. Finally,
a P-controller is used for the navigation of the robot. This enables
the robot to follow a certain path, which may be defined in the first
phase of our approach. The actual experiments with a SCITOS G5 service
robot confirm the validity and robustness of our approach.},
days = {15},
keywords = {RFID-based Navigation, Particle Filter, Fingerprints, Received Signal
Strength (RSS)},
url = {http://www.cogsys.cs.uni-tuebingen.de/publikationen/2011/ranliu2011-path-following-rss.pdf}
}