Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Rearrangement with Nonprehensile Manipulation Using Deep Reinforcement Learning
Hong Kong Univ Sci & Technol, Hong Kong, Hong Kong, Peoples R China.;HKUST Robot Inst, Hong Kong, Hong Kong, Peoples R China.;Dept Elect & Comp Engn, Hong Kong, Hong Kong, Peoples R China..
KTH, School of Electrical Engineering and Computer Science (EECS), Robotics, Perception and Learning, RPL. KTH, School of Electrical Engineering and Computer Science (EECS), Centres, Centre for Autonomous Systems, CAS.
KTH, School of Electrical Engineering and Computer Science (EECS), Robotics, Perception and Learning, RPL. KTH, School of Electrical Engineering and Computer Science (EECS), Centres, Centre for Autonomous Systems, CAS.ORCID iD: 0000-0003-2965-2953
Hong Kong Univ Sci & Technol, Hong Kong, Hong Kong, Peoples R China.;HKUST Robot Inst, Hong Kong, Hong Kong, Peoples R China.;Dept Mech & Aerosp Engn, Hong Kong, Hong Kong, Peoples R China..
Show others and affiliations
2018 (English)In: 2018 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION (ICRA), IEEE Computer Society, 2018, p. 270-277Conference paper, Published paper (Refereed)
Abstract [en]

Rearranging objects on a tabletop surface by means of nonprehensile manipulation is a task which requires skillful interaction with the physical world. Usually, this is achieved by precisely modeling physical properties of the objects, robot, and the environment for explicit planning. In contrast, as explicitly modeling the physical environment is not always feasible and involves various uncertainties, we learn a nonprehensile rearrangement strategy with deep reinforcement learning based on only visual feedback. For this, we model the task with rewards and train a deep Q-network. Our potential field-based heuristic exploration strategy reduces the amount of collisions which lead to suboptimal outcomes and we actively balance the training set to avoid bias towards poor examples. Our training process leads to quicker learning and better performance on the task as compared to uniform exploration and standard experience replay. We demonstrate empirical evidence from simulation that our method leads to a success rate of 85%, show that our system can cope with sudden changes of the environment, and compare our performance with human level performance.

Place, publisher, year, edition, pages
IEEE Computer Society, 2018. p. 270-277
Series
IEEE International Conference on Robotics and Automation ICRA, ISSN 1050-4729
National Category
Robotics
Identifiers
URN: urn:nbn:se:kth:diva-237158ISI: 000446394500028Scopus ID: 2-s2.0-85063133829ISBN: 978-1-5386-3081-5 (print)OAI: oai:DiVA.org:kth-237158DiVA, id: diva2:1258415
Conference
IEEE International Conference on Robotics and Automation (ICRA), MAY 21-25, 2018, Brisbane, AUSTRALIA
Funder
Knut and Alice Wallenberg Foundation
Note

QC 20181024

Available from: 2018-10-24 Created: 2018-10-24 Last updated: 2019-08-20Bibliographically approved

Open Access in DiVA

No full text in DiVA

Other links

Scopusconference

Authority records BETA

Stork, Johannes A.Kragic, Danica

Search in DiVA

By author/editor
Stork, Johannes A.Kragic, Danica
By organisation
Robotics, Perception and Learning, RPLCentre for Autonomous Systems, CAS
Robotics

Search outside of DiVA

GoogleGoogle Scholar

isbn
urn-nbn

Altmetric score

isbn
urn-nbn
Total: 84 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf