Advanced search
1 file | 1.07 MB

Supervised autonomy for online learning in human-robot interaction

Author
Organization
Abstract
When a robot is learning it needs to explore its environment and how its environment responds on its actions. When the environment is large and there are a large number of possible actions the robot can take, this exploration phase can take prohibitively long. However, exploration can often be optimised by letting a human expert guide the robot during its learning. Interactive machine learning, in which a human user interactively guides the robot as it learns, has been shown to be an effective way to teach a robot. It requires an intuitive control mechanism to allow the human expert to provide feedback on the robot's progress. This paper presents a novel method which combines Reinforcement Learning and Supervised Progressively Autonomous Robot Competencies (SPARC). By allowing the user to fully control the robot and by treating rewards as implicit, SPARC aims to learn an action policy while maintaining human supervisory oversight of the robot's behaviour. This method is evaluated and compared to Interactive Reinforcement Learning in a robot teaching task. Qualitative and quantitative results indicate that SPARC allows for safer and faster learning by the robot, whilst not placing a high workload on the human teacher. (C) 2017 Elsevier B.V. All rights reserved.
Keywords
IBCN, Human-Robot interaction, Reinforcement learning, Interactive machine learning, Robotics, Progressive Autonomy, Supervised autonomy

Downloads

  • (...).pdf
    • full text
    • |
    • UGent only
    • |
    • PDF
    • |
    • 1.07 MB

Citation

Please use this url to cite or link to this publication:

Chicago
Senft, Emmanuel, Paul Baxter, James Kennedy, Severin Lemaignan, and Tony Belpaeme. 2017. “Supervised Autonomy for Online Learning in Human-robot Interaction.” Pattern Recognition Letters 99: 77–86.
APA
Senft, Emmanuel, Baxter, P., Kennedy, J., Lemaignan, S., & Belpaeme, T. (2017). Supervised autonomy for online learning in human-robot interaction. PATTERN RECOGNITION LETTERS, 99, 77–86. Presented at the Workshop on Behavior Adaptation, Interaction and Learning for Assistive Robotics (BAILAR).
Vancouver
1.
Senft E, Baxter P, Kennedy J, Lemaignan S, Belpaeme T. Supervised autonomy for online learning in human-robot interaction. PATTERN RECOGNITION LETTERS. 2016; 2017;99:77–86.
MLA
Senft, Emmanuel, Paul Baxter, James Kennedy, et al. “Supervised Autonomy for Online Learning in Human-robot Interaction.” PATTERN RECOGNITION LETTERS 99 (2017): 77–86. Print.
@article{8537128,
  abstract     = {When a robot is learning it needs to explore its environment and how its environment responds on its actions. When the environment is large and there are a large number of possible actions the robot can take, this exploration phase can take prohibitively long. However, exploration can often be optimised by letting a human expert guide the robot during its learning. Interactive machine learning, in which a human user interactively guides the robot as it learns, has been shown to be an effective way to teach a robot. It requires an intuitive control mechanism to allow the human expert to provide feedback on the robot's progress. This paper presents a novel method which combines Reinforcement Learning and Supervised Progressively Autonomous Robot Competencies (SPARC). By allowing the user to fully control the robot and by treating rewards as implicit, SPARC aims to learn an action policy while maintaining human supervisory oversight of the robot's behaviour. This method is evaluated and compared to Interactive Reinforcement Learning in a robot teaching task. Qualitative and quantitative results indicate that SPARC allows for safer and faster learning by the robot, whilst not placing a high workload on the human teacher. (C) 2017 Elsevier B.V. All rights reserved.},
  author       = {Senft, Emmanuel and Baxter, Paul and Kennedy, James and Lemaignan, Severin and Belpaeme, Tony},
  issn         = {0167-8655},
  journal      = {PATTERN RECOGNITION LETTERS},
  keyword      = {IBCN,Human-Robot interaction,Reinforcement learning,Interactive machine learning,Robotics,Progressive Autonomy,Supervised autonomy},
  language     = {eng},
  location     = {Teachers College, NY},
  pages        = {77--86},
  title        = {Supervised autonomy for online learning in human-robot interaction},
  url          = {http://dx.doi.org/10.1016/j.patrec.2017.03.015},
  volume       = {99},
  year         = {2017},
}

Altmetric
View in Altmetric
Web of Science
Times cited: