Gal A. Kaminka's Publications

Sorted by DateClassified by Publication TypeClassified by TopicGrouped by Student (current)Grouped by Former Students

Area Coverage by a Multi-Robot System

Ido Ikar. Area Coverage by a Multi-Robot System. Master's Thesis, Bar Ilan University,2007.

Download

[PDF]921.1kB  

Abstract

Area coverage is an important task which can be found in many essential applications. For these applications utilizing a large number of identical and inexpensive robotic platforms, which are able to perform area coverage, may provide an appropriate solution. For example, security tasks in pubic or government buildings, usually involve human patrol, using a multi-robot system would have been good. Much work has been done in the area of multi robot coverage. Many of the existing algorithms rely on communication and advanced sensors used for maps creation and localization but problems such as communication loss, localization problems due to sensors inaccuracy and many other problems which could occur in the real world are not fully addressed. We present a new robust online algorithm for multi-robot coverage which behaves as a black box designed to work with single-robot coverage existing algorithms. Our algorithm maximizes the area coverage and minimizes the standard deviation over multiple operations for these algorithms. Our algorithm does not perform any localization and therefore requires only few common sensors which can be found in many of today's robotic platforms, this means no hardware modifications to the robots are required in order to support our algorithm. Our algorithm does not rely on communication protocols between the robots and is therefore not sensitive to any communication losses. We perform extensive experiments for our algorithm which includes different environments, different number of robots, simulated errors which might occur in the real world etc. We show that real robotic platforms can benefit from using our algorithm.

Additional Information

BibTeX

@MastersThesis{idoikar-msc, 
  author = {Ido Ikar}, 
  title = {Area Coverage by a Multi-Robot System}, 
  school = 	 {{B}ar {I}lan {U}niversity}, 
  year = 	 {2007}, 
  OPTkey = 	 {}, 
  OPTtype = 	 {}, 
  OPTaddress = 	 {}, 
  OPTmonth = 	 {}, 
  note = 	 {}, 
  abstract = { Area coverage is an important task which can be found in many essential 
applications. For these applications utilizing a large number of identical and 
inexpensive robotic platforms, which are able to perform area coverage, may 
provide an appropriate solution. For example, security tasks in pubic or 
government buildings, usually involve human patrol, using a multi-robot system 
would have been good. 
    Much work has been done in the area of multi robot coverage. Many of the 
existing algorithms rely on communication and advanced sensors used for maps 
creation and localization but problems such as communication loss, localization 
problems due to sensors inaccuracy and many other problems which could 
occur in the real world are not fully addressed. 
    We present a new robust online algorithm for multi-robot coverage which 
behaves as a black box designed to work with single-robot coverage existing 
algorithms. Our algorithm maximizes the area coverage and minimizes the 
standard deviation over multiple operations for these algorithms. Our algorithm 
does not perform any localization and therefore requires only few common 
sensors which can be found in many of today's robotic platforms, this means no 
hardware modifications to the robots are required in order to support our 
algorithm. Our algorithm does not rely on communication protocols between the 
robots and is therefore not sensitive to any communication losses. We perform 
extensive experiments for our algorithm which includes different environments, 
different number of robots, simulated errors which might occur in the real world 
etc. We show that real robotic platforms can benefit from using our algorithm. }, 
  wwwnote = {}, 
 OPTannote = 	 {} 
} 

Generated by bib2html.pl (written by Patrick Riley ) on Sun Oct 29, 2017 21:31:22