Tracking under Uncertainty with Cooperating Objects using MOMDPs

J. Capitan, L. Merino, and A. Ollero. Tracking under Uncertainty with Cooperating Objects using MOMDPs. In The First International Workshop on Networks of Cooperating Objects CONET, 2010.

Download

[PDF] 

Abstract

Planning under uncertainty faces a scalability problem whenconsidering multi-robot teams, as the information spacescales exponentially with the number of robots. To addressthis issue, this paper proposes to decentralize multi-agentPartially Observable Markov Decision Process (POMDPs)while maintaining cooperation between robots by usingPOMDP policy auctions. Auctions provide a flexible wayof coordinating individual policies modeled by POMDPsand have low communication requirements. Additionally,communication models in the multi-agent POMDP literatureseverely mismatch with real inter-robot communication.We address this issue by applying a decentralized data fusionmethod in order to efficiently maintain a joint beliefstate among the robots. The paper focuses on a cooperativetracking application, in which several robots have tojointly track a moving target of interest. The proposed ideasare illustrated in real multi-robot experiments, showcasingthe flexible and robust coordination that our techniques canprovide

BibTeX Entry

@INPROCEEDINGS{capitan_conet10,
  author = {Capitan, J. and Merino, L. and Ollero, A.},
  title = {{T}racking under {U}ncertainty with {C}ooperating {O}bjects using
	{MOMDP}s},
  booktitle = {{T}he {F}irst {I}nternational {W}orkshop on {N}etworks of {C}ooperating
	{O}bjects {CONET}},
  year = {2010},
  abstract={Planning under uncertainty faces a scalability problem when
considering multi-robot teams, as the information space
scales exponentially with the number of robots. To address
this issue, this paper proposes to decentralize multi-agent
Partially Observable Markov Decision Process (POMDPs)
while maintaining cooperation between robots by using
POMDP policy auctions. Auctions provide a flexible way
of coordinating individual policies modeled by POMDPs
and have low communication requirements. Additionally,
communication models in the multi-agent POMDP literature
severely mismatch with real inter-robot communication.
We address this issue by applying a decentralized data fusion
method in order to efficiently maintain a joint belief
state among the robots. The paper focuses on a cooperative
tracking application, in which several robots have to
jointly track a moving target of interest. The proposed ideas
are illustrated in real multi-robot experiments, showcasing
the flexible and robust coordination that our techniques can
provide},
}

Generated by bib2html.pl (written by Patrick Riley ) on Thu Aug 16, 2018 19:28:37