Communication-Aware Multi-Agent Metareasoning for Decentralized Task Allocation

<italic>Metareasoning</italic> refers to reasoning about one’s own decision making process. This paper considers metareasoning about the decision making process in multi-agent settings. We present a multi-agent metareasoning approach that enables a multi-agent team to select w...

Full description

Bibliographic Details
Main Authors:	Estefany Carrillo, Suyash Yeotikar, Sharan Nayak, Mohamed Khalid M. Jaffar, Shapour Azarm, Jeffrey W. Herrmann, Michael Otte, Huan Xu
Format:	Article
Language:	English
Published:	IEEE 2021-01-01
Series:	IEEE Access
Subjects:	Distributed robot systems decentralized task allocation meta-level control
Online Access:	https://ieeexplore.ieee.org/document/9481127/

id	doaj-199f10a49496483c912fee2b42cfc179
record_format	Article
spelling	doaj-199f10a49496483c912fee2b42cfc1792021-07-19T23:00:25ZengIEEEIEEE Access2169-35362021-01-019987129873010.1109/ACCESS.2021.30962299481127Communication-Aware Multi-Agent Metareasoning for Decentralized Task AllocationEstefany Carrillo0https://orcid.org/0000-0001-5062-6923Suyash Yeotikar1Sharan Nayak2https://orcid.org/0000-0002-1191-7792Mohamed Khalid M. Jaffar3Shapour Azarm4https://orcid.org/0000-0001-5248-6266Jeffrey W. Herrmann5https://orcid.org/0000-0002-4081-1196Michael Otte6https://orcid.org/0000-0001-7432-0734Huan Xu7https://orcid.org/0000-0002-7238-8759Department of Aerospace Engineering, University of Maryland at College Park, College Park, MD, USADepartment of Aerospace Engineering, University of Maryland at College Park, College Park, MD, USADepartment of Aerospace Engineering, University of Maryland at College Park, College Park, MD, USADepartment of Aerospace Engineering, University of Maryland at College Park, College Park, MD, USADepartment of Mechanical Engineering, University of Maryland at College Park, College Park, MD, USADepartment of Mechanical Engineering, University of Maryland at College Park, College Park, MD, USADepartment of Aerospace Engineering, University of Maryland at College Park, College Park, MD, USADepartment of Aerospace Engineering, University of Maryland at College Park, College Park, MD, USA<italic>Metareasoning</italic> refers to reasoning about one’s own decision making process. This paper considers metareasoning about the decision making process in multi-agent settings. We present a multi-agent metareasoning approach that enables a multi-agent team to select which task allocation algorithm to use as a function of changing communication quality level. Given a set of multi-agent task allocation algorithms, we synthesize a policy that prescribes the best algorithm to use among a predefined set of algorithms for a given communication level. Since each agent in the team runs the same policy, the team (or a part of the team) will collectively switch between task allocation algorithms as a function of the observed level of communication. We apply reactive synthesis to generate the policy from high-level specifications written in Linear Temporal Logic encoding the agents’ switching behavior with respect to the state of the environment. We perform experiments in simulation to identify the best performing algorithms under different communication levels. The communication environment is modeled using the Rayleigh fading model and communication estimation is done through the exchange of heartbeat messages among agents. We test our metareasoning policy in three types of scenarios: search & rescue, fire monitoring, and ship protection scenarios. For each scenario, we demonstrate that our policy achieved better performance with respect to either max distance traveled, max number of transmitted messages or both compared to running any single algorithm.https://ieeexplore.ieee.org/document/9481127/Distributed robot systemsdecentralized task allocationmeta-level control
collection	DOAJ
language	English
format	Article
sources	DOAJ
author	Estefany Carrillo Suyash Yeotikar Sharan Nayak Mohamed Khalid M. Jaffar Shapour Azarm Jeffrey W. Herrmann Michael Otte Huan Xu
spellingShingle	Estefany Carrillo Suyash Yeotikar Sharan Nayak Mohamed Khalid M. Jaffar Shapour Azarm Jeffrey W. Herrmann Michael Otte Huan Xu Communication-Aware Multi-Agent Metareasoning for Decentralized Task Allocation IEEE Access Distributed robot systems decentralized task allocation meta-level control
author_facet	Estefany Carrillo Suyash Yeotikar Sharan Nayak Mohamed Khalid M. Jaffar Shapour Azarm Jeffrey W. Herrmann Michael Otte Huan Xu
author_sort	Estefany Carrillo
title	Communication-Aware Multi-Agent Metareasoning for Decentralized Task Allocation
title_short	Communication-Aware Multi-Agent Metareasoning for Decentralized Task Allocation
title_full	Communication-Aware Multi-Agent Metareasoning for Decentralized Task Allocation
title_fullStr	Communication-Aware Multi-Agent Metareasoning for Decentralized Task Allocation
title_full_unstemmed	Communication-Aware Multi-Agent Metareasoning for Decentralized Task Allocation
title_sort	communication-aware multi-agent metareasoning for decentralized task allocation
publisher	IEEE
series	IEEE Access
issn	2169-3536
publishDate	2021-01-01
description	<italic>Metareasoning</italic> refers to reasoning about one’s own decision making process. This paper considers metareasoning about the decision making process in multi-agent settings. We present a multi-agent metareasoning approach that enables a multi-agent team to select which task allocation algorithm to use as a function of changing communication quality level. Given a set of multi-agent task allocation algorithms, we synthesize a policy that prescribes the best algorithm to use among a predefined set of algorithms for a given communication level. Since each agent in the team runs the same policy, the team (or a part of the team) will collectively switch between task allocation algorithms as a function of the observed level of communication. We apply reactive synthesis to generate the policy from high-level specifications written in Linear Temporal Logic encoding the agents’ switching behavior with respect to the state of the environment. We perform experiments in simulation to identify the best performing algorithms under different communication levels. The communication environment is modeled using the Rayleigh fading model and communication estimation is done through the exchange of heartbeat messages among agents. We test our metareasoning policy in three types of scenarios: search & rescue, fire monitoring, and ship protection scenarios. For each scenario, we demonstrate that our policy achieved better performance with respect to either max distance traveled, max number of transmitted messages or both compared to running any single algorithm.
topic	Distributed robot systems decentralized task allocation meta-level control
url	https://ieeexplore.ieee.org/document/9481127/
work_keys_str_mv	AT estefanycarrillo communicationawaremultiagentmetareasoningfordecentralizedtaskallocation AT suyashyeotikar communicationawaremultiagentmetareasoningfordecentralizedtaskallocation AT sharannayak communicationawaremultiagentmetareasoningfordecentralizedtaskallocation AT mohamedkhalidmjaffar communicationawaremultiagentmetareasoningfordecentralizedtaskallocation AT shapourazarm communicationawaremultiagentmetareasoningfordecentralizedtaskallocation AT jeffreywherrmann communicationawaremultiagentmetareasoningfordecentralizedtaskallocation AT michaelotte communicationawaremultiagentmetareasoningfordecentralizedtaskallocation AT huanxu communicationawaremultiagentmetareasoningfordecentralizedtaskallocation
_version_	1721294369892335616

Communication-Aware Multi-Agent Metareasoning for Decentralized Task Allocation

Similar Items