HactarV2: An agent team strategy based on implicit coordination

Marc Dekker*, Pieter Hameete, Michiel Hegemans, Sebastiaan Leysen, Joris Van Den Oever, Jeff Smits, Koen V. Hindriks

*Corresponding author for this work

Research output: Chapter in Book / Report / Conference proceedingConference contributionAcademicpeer-review

Abstract

In this paper we report on the design and implementation of our multi-agent system, called HactarV2, for the Agent Contest 2011. HactarV2 has been implemented in the agent programming language Goal. One of the main challenges of the Agent Contest is to design a decentralized multi-agent system that is able to strategically compete with other agent teams. To address this challenge, the strategy of HactarV2 is based on implicit coordination between agents and there is no central manager that keeps track of all information. The aim, moreover, has been to minimize the communication between agents. Communication is used by HactarV2 agents to ensure that each of them maintains the same map of the environment. The Mars scenario of this year required agents to explore, locate and occupy high valued zones on the planet Mars. Because initially agents are randomly placed on the map, in the first phase of the game the agents individually explore the map and update each other. Agents have different roles and we describe the strategies used by individual agents per role. In the second phase of the game, which starts when the agents have located high value nodes on the map, the agents group together and act as a swarm to maintain and possibly expand the zone on the map that is occupied by the agents.

Original languageEnglish
Title of host publicationProgramming Multi-Agent Systems - 9th International Workshop, ProMAS 2011, Revised Selected Papers
Pages173-184
Number of pages12
DOIs
Publication statusPublished - 15 Aug 2012
Externally publishedYes
Event9th International Workshop on ProgrammingMulti-Agent Systems, ProMAS 2011 - Taipei, Taiwan, Province of China
Duration: 3 May 20113 May 2011

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume7217 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference9th International Workshop on ProgrammingMulti-Agent Systems, ProMAS 2011
CountryTaiwan, Province of China
CityTaipei
Period3/05/113/05/11

    Fingerprint

Cite this

Dekker, M., Hameete, P., Hegemans, M., Leysen, S., Van Den Oever, J., Smits, J., & Hindriks, K. V. (2012). HactarV2: An agent team strategy based on implicit coordination. In Programming Multi-Agent Systems - 9th International Workshop, ProMAS 2011, Revised Selected Papers (pp. 173-184). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7217 LNAI). https://doi.org/10.1007/978-3-642-31915-0_10