Meta-strategy for cooperative tasks with learning of environments in multi-agent continuous tasks

Ayumi Sugiyama, Toshiharu Sugawara

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    5 Citations (Scopus)

    Abstract

    With the development of robot technology, we can expect selfpropelled robots working in large areas where cooperative and coordinated behaviors by multiple (hardware and software) robots are necessary. However, it is not trivial for agents, which are control programs running on robots, to determine the actions for their cooperative behaviors, because such strategies depend on the characteristics of the environment and the capabilities of individual agents. Therefore, using the example of continuous cleaning tasks by multiple agents, we propose a method of meta-strategy that decide the appropriate planning strategies for cooperation and coordination through with the learning of the performance of individual strategies and the environmental data in a multi-agent systems context, but without complex reasoning for deep coordination due to the limited CPU capability and battery capacity. We experimentally evaluated our method by comparing it with a conventional method that assumes that agents have knowledge on where agents visit frequently (since they are easy to become dirty). We found that agents with the proposed method could operate as effectively as and, in complex areas, outperformed those with the conventional method. Finally, we describe that the reasons for such a counterintuitive phenomenon is induced from splitting up in working by autonomous agents based on the local observations. We also discuss the limitation of the current method.

    Original languageEnglish
    Title of host publicationProceedings of the ACM Symposium on Applied Computing
    PublisherAssociation for Computing Machinery
    Pages494-500
    Number of pages7
    Volume13-17-April-2015
    ISBN (Print)9781450331968
    DOIs
    Publication statusPublished - 2015 Apr 13
    Event30th Annual ACM Symposium on Applied Computing, SAC 2015 - Salamanca, Spain
    Duration: 2015 Apr 132015 Apr 17

    Other

    Other30th Annual ACM Symposium on Applied Computing, SAC 2015
    CountrySpain
    CitySalamanca
    Period15/4/1315/4/17

    Fingerprint

    Robots
    Autonomous agents
    Multi agent systems
    Computer hardware
    Program processors
    Cleaning
    Planning

    Keywords

    • Continuous cleaning
    • Cooperation
    • Coordination
    • Division of labor
    • Multi-agent systems

    ASJC Scopus subject areas

    • Software

    Cite this

    Sugiyama, A., & Sugawara, T. (2015). Meta-strategy for cooperative tasks with learning of environments in multi-agent continuous tasks. In Proceedings of the ACM Symposium on Applied Computing (Vol. 13-17-April-2015, pp. 494-500). Association for Computing Machinery. https://doi.org/10.1145/2695664.2695878

    Meta-strategy for cooperative tasks with learning of environments in multi-agent continuous tasks. / Sugiyama, Ayumi; Sugawara, Toshiharu.

    Proceedings of the ACM Symposium on Applied Computing. Vol. 13-17-April-2015 Association for Computing Machinery, 2015. p. 494-500.

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Sugiyama, A & Sugawara, T 2015, Meta-strategy for cooperative tasks with learning of environments in multi-agent continuous tasks. in Proceedings of the ACM Symposium on Applied Computing. vol. 13-17-April-2015, Association for Computing Machinery, pp. 494-500, 30th Annual ACM Symposium on Applied Computing, SAC 2015, Salamanca, Spain, 15/4/13. https://doi.org/10.1145/2695664.2695878
    Sugiyama A, Sugawara T. Meta-strategy for cooperative tasks with learning of environments in multi-agent continuous tasks. In Proceedings of the ACM Symposium on Applied Computing. Vol. 13-17-April-2015. Association for Computing Machinery. 2015. p. 494-500 https://doi.org/10.1145/2695664.2695878
    Sugiyama, Ayumi ; Sugawara, Toshiharu. / Meta-strategy for cooperative tasks with learning of environments in multi-agent continuous tasks. Proceedings of the ACM Symposium on Applied Computing. Vol. 13-17-April-2015 Association for Computing Machinery, 2015. pp. 494-500
    @inproceedings{b7134bd393a64158998504c502cf10d9,
    title = "Meta-strategy for cooperative tasks with learning of environments in multi-agent continuous tasks",
    abstract = "With the development of robot technology, we can expect selfpropelled robots working in large areas where cooperative and coordinated behaviors by multiple (hardware and software) robots are necessary. However, it is not trivial for agents, which are control programs running on robots, to determine the actions for their cooperative behaviors, because such strategies depend on the characteristics of the environment and the capabilities of individual agents. Therefore, using the example of continuous cleaning tasks by multiple agents, we propose a method of meta-strategy that decide the appropriate planning strategies for cooperation and coordination through with the learning of the performance of individual strategies and the environmental data in a multi-agent systems context, but without complex reasoning for deep coordination due to the limited CPU capability and battery capacity. We experimentally evaluated our method by comparing it with a conventional method that assumes that agents have knowledge on where agents visit frequently (since they are easy to become dirty). We found that agents with the proposed method could operate as effectively as and, in complex areas, outperformed those with the conventional method. Finally, we describe that the reasons for such a counterintuitive phenomenon is induced from splitting up in working by autonomous agents based on the local observations. We also discuss the limitation of the current method.",
    keywords = "Continuous cleaning, Cooperation, Coordination, Division of labor, Multi-agent systems",
    author = "Ayumi Sugiyama and Toshiharu Sugawara",
    year = "2015",
    month = "4",
    day = "13",
    doi = "10.1145/2695664.2695878",
    language = "English",
    isbn = "9781450331968",
    volume = "13-17-April-2015",
    pages = "494--500",
    booktitle = "Proceedings of the ACM Symposium on Applied Computing",
    publisher = "Association for Computing Machinery",

    }

    TY - GEN

    T1 - Meta-strategy for cooperative tasks with learning of environments in multi-agent continuous tasks

    AU - Sugiyama, Ayumi

    AU - Sugawara, Toshiharu

    PY - 2015/4/13

    Y1 - 2015/4/13

    N2 - With the development of robot technology, we can expect selfpropelled robots working in large areas where cooperative and coordinated behaviors by multiple (hardware and software) robots are necessary. However, it is not trivial for agents, which are control programs running on robots, to determine the actions for their cooperative behaviors, because such strategies depend on the characteristics of the environment and the capabilities of individual agents. Therefore, using the example of continuous cleaning tasks by multiple agents, we propose a method of meta-strategy that decide the appropriate planning strategies for cooperation and coordination through with the learning of the performance of individual strategies and the environmental data in a multi-agent systems context, but without complex reasoning for deep coordination due to the limited CPU capability and battery capacity. We experimentally evaluated our method by comparing it with a conventional method that assumes that agents have knowledge on where agents visit frequently (since they are easy to become dirty). We found that agents with the proposed method could operate as effectively as and, in complex areas, outperformed those with the conventional method. Finally, we describe that the reasons for such a counterintuitive phenomenon is induced from splitting up in working by autonomous agents based on the local observations. We also discuss the limitation of the current method.

    AB - With the development of robot technology, we can expect selfpropelled robots working in large areas where cooperative and coordinated behaviors by multiple (hardware and software) robots are necessary. However, it is not trivial for agents, which are control programs running on robots, to determine the actions for their cooperative behaviors, because such strategies depend on the characteristics of the environment and the capabilities of individual agents. Therefore, using the example of continuous cleaning tasks by multiple agents, we propose a method of meta-strategy that decide the appropriate planning strategies for cooperation and coordination through with the learning of the performance of individual strategies and the environmental data in a multi-agent systems context, but without complex reasoning for deep coordination due to the limited CPU capability and battery capacity. We experimentally evaluated our method by comparing it with a conventional method that assumes that agents have knowledge on where agents visit frequently (since they are easy to become dirty). We found that agents with the proposed method could operate as effectively as and, in complex areas, outperformed those with the conventional method. Finally, we describe that the reasons for such a counterintuitive phenomenon is induced from splitting up in working by autonomous agents based on the local observations. We also discuss the limitation of the current method.

    KW - Continuous cleaning

    KW - Cooperation

    KW - Coordination

    KW - Division of labor

    KW - Multi-agent systems

    UR - http://www.scopus.com/inward/record.url?scp=84955507435&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84955507435&partnerID=8YFLogxK

    U2 - 10.1145/2695664.2695878

    DO - 10.1145/2695664.2695878

    M3 - Conference contribution

    AN - SCOPUS:84955507435

    SN - 9781450331968

    VL - 13-17-April-2015

    SP - 494

    EP - 500

    BT - Proceedings of the ACM Symposium on Applied Computing

    PB - Association for Computing Machinery

    ER -