Publications

Sorted by DateClassified by Publication TypeClassified by Research Category

Decentralised Online Planning for Multi-Robot Warehouse Commisioning

Daniel Claes, Frans A. Oliehoek, Hendrik Baier, and Karl Tuyls. Decentralised Online Planning for Multi-Robot Warehouse Commisioning. In Proceedings of the Sixteenth International Conference on Autonomous Agents and Multiagent Systems, pp. 492–500, May 2017.

Download

pdf [798.5kB]  

Abstract

Warehouse commissioning is a complex task in which a teamof robots needs to gather and deliver items as fast and effi-ciently as possible while adhering to the constraint capacityof the robots. Typical centralised control approaches canquickly become infeasible when dealing with many robots.Instead, we tackle this spatial task allocation problem viadistributed planning on each robot in the system. Stateof the art distributed planning approaches suffer from anumber of limiting assumptions and ad-hoc approximations.This paper demonstrates how to use Monte Carlo Tree Search(MCTS) to overcome these limitations and provide scalabil-ity in a more principled manner. Our simulation-based eval-uation demonstrates that this translates to higher task per-formance, especially when tasks get more complex. More-over, this higher performance does not come at the cost ofscalability: in fact, the proposed approach scales better thanthe previous best approach, demonstrating excellent perfor-mance on an 8-robot team servicing a warehouse comprisedof over 200 locations.

BibTeX Entry

@inproceedings{Claes17AAMAS,
    author    = {Daniel Claes  and  Frans A. Oliehoek and
                 Hendrik Baier and Karl Tuyls},
    title =     {Decentralised Online Planning for Multi-Robot Warehouse Commisioning},
    booktitle = AAMAS17,
    year      = {2017},
    month =     may,
    pages =     {492--500},
    abstract = {
Warehouse commissioning is a complex task in which a team
of robots needs to gather and deliver items as fast and effi-
ciently as possible while adhering to the constraint capacity
of the robots. Typical centralised control approaches can
quickly become infeasible when dealing with many robots.
Instead, we tackle this spatial task allocation problem via
distributed planning on each robot in the system. State
of the art distributed planning approaches suffer from a
number of limiting assumptions and ad-hoc approximations.
This paper demonstrates how to use Monte Carlo Tree Search
(MCTS) to overcome these limitations and provide scalabil-
ity in a more principled manner. Our simulation-based eval-
uation demonstrates that this translates to higher task per-
formance, especially when tasks get more complex. More-
over, this higher performance does not come at the cost of
scalability: in fact, the proposed approach scales better than
the previous best approach, demonstrating excellent perfor-
mance on an 8-robot team servicing a warehouse comprised
of over 200 locations.
    }
}

Generated by bib2html.pl (written by Patrick Riley) on Fri Sep 15, 2017 12:26:00 UTC