Hierarchical learning of robotic contact policies

Simonič, Mihael (author), Ude, Aleš (author), Nemec, Bojan (author)

The paper addresses the issue of learning tasks where a robot maintains permanent contact with the environment. We propose a new methodology based on a hierarchical learning scheme coupled with task representation through directed graphs. These graphs are constituted of nodes and branches that correspond to the states and robotic actions, respectively. The upper level of the hierarchy essentially operates as a decision-making algorithm. It leverages reinforcement learning (RL) techniques to facilitate optimal decision-making. The actions are generated by a constraint-space following (CSF) controller that autonomously identifies feasible directions for motion. The controller generates robot motion by adjusting its stiffness in the direction defined by the Frenet–Serret frame, which is aligned with the robot path. The proposed framework was experimentally verified through a series of challenging robotic tasks such as maze learning, door opening, learning to shift the manual car gear, and learning car license plate light assembly by disassembly.

• Language: English
• Keywords: autonomous robot learning, learning, experience, compliance and impedance control, learning from experience, perception-action coupling, compliant assembly
• Work type: Article
• Typology: 1.01 - Original Scientific Article
• Organization: FE - Faculty of Electrical Engineering
• Publication status: Published
• Publication version: Version of Record
• Year: 2024
• Number of pages: 12 str.
• Numbering: Vol. 86, art. 102657
• PID: 20.500.12556/RUL-153103
• UDC: 004
• ISSN on article: 1879-2537
• DOI: 10.1016/j.rcim.2023.102657
• COBISS.SI-ID: 164894211
• Publication date in RUL: 18.12.2023

Record is a part of a journal

• Title: Robotics and computer-integrated manufacturing
• Publisher: Elsevier
• ISSN: 1879-2537
• COBISS.SI-ID: 18810902

Licences

• License: CC BY-NC 4.0, Creative Commons Attribution-NonCommercial 4.0 International
• Link: http://creativecommons.org/licenses/by-nc/4.0/
• Description: A creative commons license that bans commercial use, but the users don’t have to license their derivative works on the same terms.

Secondary language

• Language: Slovenian
• Keywords: strojno učenje, učenje, izkušnje

Projects

• Funder: EC - European Commission
• Funding programme: H2020
• Project number: 871352
• Name: Self-reconfiguration of a robotic workcell for the recycling of electronic waste
• Acronym: ReconCycle

• Funder: ARRS - Slovenian Research Agency
• Project number: P2-0076
• Name: Avtomatika, robotika in biokibernetika