Publikation:

Revealing principles of autonomous thermal soaring in windy conditions using vulture-inspired deep reinforcement-learning

Lade...
Vorschaubild

Dateien

Flato_2-1b3gogojjl8x3.pdf
Flato_2-1b3gogojjl8x3.pdfGröße: 4.94 MBDownloads: 26

Datum

2024

Autor:innen

Flato, Yoav
Tamar, Aviv
Nathan, Ran
Beatus, Tsevi

Herausgeber:innen

Kontakt

ISSN der Zeitschrift

Electronic ISSN

ISBN

Bibliografische Daten

Verlag

Schriftenreihe

Auflagebezeichnung

ArXiv-ID

Internationale Patentnummer

Link zur Lizenz

Angaben zur Forschungsförderung

Projekt

Open Access-Veröffentlichung
Open Access Gold
Core Facility der Universität Konstanz

Gesperrt bis

Titel in einer weiteren Sprache

Publikationstyp
Zeitschriftenartikel
Publikationsstatus
Published

Erschienen in

Nature Communications. Springer. 2024, 15, 4942. eISSN 2041-1723. Verfügbar unter: doi: 10.1038/s41467-024-48670-x

Zusammenfassung

Thermal soaring, a technique used by birds and gliders to utilize updrafts of hot air, is an appealing model-problem for studying motion control and how it is learned by animals and engineered autonomous systems. Thermal soaring has rich dynamics and nontrivial constraints, yet it uses few control parameters and is becoming experimentally accessible. Following recent developments in applying reinforcement learning methods for training deep neural-network (deep-RL) models to soar autonomously both in simulation and real gliders, here we develop a simulation-based deep-RL system to study the learning process of thermal soaring. We find that this process has learning bottlenecks, we define a new efficiency metric and use it to characterize learning robustness, we compare the learned policy to data from soaring vultures, and find that the neurons of the trained network divide into function clusters that evolve during learning. These results pose thermal soaring as a rich yet tractable model-problem for the learning of motion control.

Zusammenfassung in einer weiteren Sprache

Fachgebiet (DDC)
570 Biowissenschaften, Biologie

Schlagwörter

Konferenz

Rezension
undefined / . - undefined, undefined

Forschungsvorhaben

Organisationseinheiten

Zeitschriftenheft

Zugehörige Datensätze in KOPS

Zitieren

ISO 690FLATO, Yoav, Roi HAREL, Aviv TAMAR, Ran NATHAN, Tsevi BEATUS, 2024. Revealing principles of autonomous thermal soaring in windy conditions using vulture-inspired deep reinforcement-learning. In: Nature Communications. Springer. 2024, 15, 4942. eISSN 2041-1723. Verfügbar unter: doi: 10.1038/s41467-024-48670-x
BibTex
@article{Flato2024-06-10Revea-70279,
  year={2024},
  doi={10.1038/s41467-024-48670-x},
  title={Revealing principles of autonomous thermal soaring in windy conditions using vulture-inspired deep reinforcement-learning},
  volume={15},
  journal={Nature Communications},
  author={Flato, Yoav and Harel, Roi and Tamar, Aviv and Nathan, Ran and Beatus, Tsevi},
  note={Article Number: 4942}
}
RDF
<rdf:RDF
    xmlns:dcterms="http://purl.org/dc/terms/"
    xmlns:dc="http://purl.org/dc/elements/1.1/"
    xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
    xmlns:bibo="http://purl.org/ontology/bibo/"
    xmlns:dspace="http://digital-repositories.org/ontologies/dspace/0.1.0#"
    xmlns:foaf="http://xmlns.com/foaf/0.1/"
    xmlns:void="http://rdfs.org/ns/void#"
    xmlns:xsd="http://www.w3.org/2001/XMLSchema#" > 
  <rdf:Description rdf:about="https://kops.uni-konstanz.de/server/rdf/resource/123456789/70279">
    <dcterms:rights rdf:resource="http://creativecommons.org/licenses/by/4.0/"/>
    <dcterms:abstract>Thermal soaring, a technique used by birds and gliders to utilize updrafts of hot air, is an appealing model-problem for studying motion control and how it is learned by animals and engineered autonomous systems. Thermal soaring has rich dynamics and nontrivial constraints, yet it uses few control parameters and is becoming experimentally accessible. Following recent developments in applying reinforcement learning methods for training deep neural-network (deep-RL) models to soar autonomously both in simulation and real gliders, here we develop a simulation-based deep-RL system to study the learning process of thermal soaring. We find that this process has learning bottlenecks, we define a new efficiency metric and use it to characterize learning robustness, we compare the learned policy to data from soaring vultures, and find that the neurons of the trained network divide into function clusters that evolve during learning. These results pose thermal soaring as a rich yet tractable model-problem for the learning of motion control.</dcterms:abstract>
    <dc:creator>Beatus, Tsevi</dc:creator>
    <dspace:isPartOfCollection rdf:resource="https://kops.uni-konstanz.de/server/rdf/resource/123456789/43615"/>
    <foaf:homepage rdf:resource="http://localhost:8080/"/>
    <dc:creator>Harel, Roi</dc:creator>
    <dc:contributor>Beatus, Tsevi</dc:contributor>
    <dcterms:isPartOf rdf:resource="https://kops.uni-konstanz.de/server/rdf/resource/123456789/28"/>
    <dcterms:title>Revealing principles of autonomous thermal soaring in windy conditions using vulture-inspired deep reinforcement-learning</dcterms:title>
    <dc:rights>Attribution 4.0 International</dc:rights>
    <dc:language>eng</dc:language>
    <dcterms:hasPart rdf:resource="https://kops.uni-konstanz.de/bitstream/123456789/70279/1/Flato_2-1b3gogojjl8x3.pdf"/>
    <dspace:isPartOfCollection rdf:resource="https://kops.uni-konstanz.de/server/rdf/resource/123456789/28"/>
    <dc:date rdf:datatype="http://www.w3.org/2001/XMLSchema#dateTime">2024-06-28T10:56:55Z</dc:date>
    <dc:contributor>Flato, Yoav</dc:contributor>
    <dc:creator>Flato, Yoav</dc:creator>
    <bibo:uri rdf:resource="https://kops.uni-konstanz.de/handle/123456789/70279"/>
    <dc:creator>Nathan, Ran</dc:creator>
    <dc:contributor>Nathan, Ran</dc:contributor>
    <dcterms:available rdf:datatype="http://www.w3.org/2001/XMLSchema#dateTime">2024-06-28T10:56:55Z</dcterms:available>
    <dc:contributor>Tamar, Aviv</dc:contributor>
    <void:sparqlEndpoint rdf:resource="http://localhost/fuseki/dspace/sparql"/>
    <dcterms:issued>2024-06-10</dcterms:issued>
    <dcterms:isPartOf rdf:resource="https://kops.uni-konstanz.de/server/rdf/resource/123456789/43615"/>
    <dc:creator>Tamar, Aviv</dc:creator>
    <dc:contributor>Harel, Roi</dc:contributor>
    <dspace:hasBitstream rdf:resource="https://kops.uni-konstanz.de/bitstream/123456789/70279/1/Flato_2-1b3gogojjl8x3.pdf"/>
  </rdf:Description>
</rdf:RDF>

Interner Vermerk

xmlui.Submission.submit.DescribeStep.inputForms.label.kops_note_fromSubmitter

Kontakt
URL der Originalveröffentl.

Prüfdatum der URL

Prüfungsdatum der Dissertation

Finanzierungsart

Kommentar zur Publikation

Allianzlizenz
Corresponding Authors der Uni Konstanz vorhanden
Internationale Co-Autor:innen
Universitätsbibliographie
Nein
Begutachtet
Ja
Diese Publikation teilen