Rearrangement with Nonprehensile Manipulation Using Deep Reinforcement LearningVisa övriga samt affilieringar
2018 (Engelska)Ingår i: 2018 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION (ICRA), IEEE Computer Society, 2018, s. 270-277Konferensbidrag, Publicerat paper (Refereegranskat)
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.
Ort, förlag, år, upplaga, sidor
IEEE Computer Society, 2018. s. 270-277
Serie
IEEE International Conference on Robotics and Automation ICRA, ISSN 1050-4729
Nationell ämneskategori
Robotteknik och automation
Identifikatorer
URN: urn:nbn:se:kth:diva-237158ISI: 000446394500028Scopus ID: 2-s2.0-85063133829ISBN: 978-1-5386-3081-5 (tryckt)OAI: oai:DiVA.org:kth-237158DiVA, id: diva2:1258415
Konferens
IEEE International Conference on Robotics and Automation (ICRA), MAY 21-25, 2018, Brisbane, AUSTRALIA
Forskningsfinansiär
Knut och Alice Wallenbergs Stiftelse
Anmärkning
QC 20181024
2018-10-242018-10-242019-08-20Bibliografiskt granskad