Autonomous Skills Creation and Integration in Robotics

Lorenzo Riano, TM McGinnity

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

    2 Citations (Scopus)

    Abstract

    The fragmentation of research in AI and robotics has created a vast repertoire of skills a robot could be equipped with but that hardly integrate to form a complex action. We propose a novel evolutionary algorithm that aims at autonomously integrate, adapt and create new actions by re-using skills that are either externally provided or previously generated. Complex actions are created by instantiating a Finite State Automaton and new skills are created using fully recurrent neural networks.We validated our approach in two scenarios, i.e. exploration and moving to pre-grasp positions. Our experiments show that complex actions can be created by composing independently developed skills. The results have been applied and tested with a real robot in a variety of scenarios.
    LanguageEnglish
    Title of host publicationUnknown Host Publication
    Number of pages6
    Publication statusPublished - 22 Mar 2012
    EventProc. of AAAI 2012 Spring Symposium on "Designing Intelligent Robots: Reintegrating AI" 2012 - Stanford
    Duration: 22 Mar 2012 → …

    Conference

    ConferenceProc. of AAAI 2012 Spring Symposium on "Designing Intelligent Robots: Reintegrating AI" 2012
    Period22/03/12 → …

    Fingerprint

    Robotics
    Robots
    Recurrent neural networks
    Finite automata
    Evolutionary algorithms
    Experiments

    Cite this

    Riano, L., & McGinnity, TM. (2012). Autonomous Skills Creation and Integration in Robotics. In Unknown Host Publication
    Riano, Lorenzo ; McGinnity, TM. / Autonomous Skills Creation and Integration in Robotics. Unknown Host Publication. 2012.
    @inproceedings{30bee676b2ca4ff9b29639def8eed58c,
    title = "Autonomous Skills Creation and Integration in Robotics",
    abstract = "The fragmentation of research in AI and robotics has created a vast repertoire of skills a robot could be equipped with but that hardly integrate to form a complex action. We propose a novel evolutionary algorithm that aims at autonomously integrate, adapt and create new actions by re-using skills that are either externally provided or previously generated. Complex actions are created by instantiating a Finite State Automaton and new skills are created using fully recurrent neural networks.We validated our approach in two scenarios, i.e. exploration and moving to pre-grasp positions. Our experiments show that complex actions can be created by composing independently developed skills. The results have been applied and tested with a real robot in a variety of scenarios.",
    author = "Lorenzo Riano and TM McGinnity",
    year = "2012",
    month = "3",
    day = "22",
    language = "English",
    booktitle = "Unknown Host Publication",

    }

    Riano, L & McGinnity, TM 2012, Autonomous Skills Creation and Integration in Robotics. in Unknown Host Publication. Proc. of AAAI 2012 Spring Symposium on "Designing Intelligent Robots: Reintegrating AI" 2012, 22/03/12.

    Autonomous Skills Creation and Integration in Robotics. / Riano, Lorenzo; McGinnity, TM.

    Unknown Host Publication. 2012.

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

    TY - GEN

    T1 - Autonomous Skills Creation and Integration in Robotics

    AU - Riano, Lorenzo

    AU - McGinnity, TM

    PY - 2012/3/22

    Y1 - 2012/3/22

    N2 - The fragmentation of research in AI and robotics has created a vast repertoire of skills a robot could be equipped with but that hardly integrate to form a complex action. We propose a novel evolutionary algorithm that aims at autonomously integrate, adapt and create new actions by re-using skills that are either externally provided or previously generated. Complex actions are created by instantiating a Finite State Automaton and new skills are created using fully recurrent neural networks.We validated our approach in two scenarios, i.e. exploration and moving to pre-grasp positions. Our experiments show that complex actions can be created by composing independently developed skills. The results have been applied and tested with a real robot in a variety of scenarios.

    AB - The fragmentation of research in AI and robotics has created a vast repertoire of skills a robot could be equipped with but that hardly integrate to form a complex action. We propose a novel evolutionary algorithm that aims at autonomously integrate, adapt and create new actions by re-using skills that are either externally provided or previously generated. Complex actions are created by instantiating a Finite State Automaton and new skills are created using fully recurrent neural networks.We validated our approach in two scenarios, i.e. exploration and moving to pre-grasp positions. Our experiments show that complex actions can be created by composing independently developed skills. The results have been applied and tested with a real robot in a variety of scenarios.

    M3 - Conference contribution

    BT - Unknown Host Publication

    ER -