Autonomous Motion
Note: This department has relocated.

Skill learning and task outcome prediction for manipulation

2011

Conference Paper

am

Learning complex motor skills for real world tasks is a hard problem in robotic manipulation that often requires painstaking manual tuning and design by a human expert. In this work, we present a Reinforcement Learning based approach to acquiring new motor skills from demonstration. Our approach allows the robot to learn fine manipulation skills and significantly improve its success rate and skill level starting from a possibly coarse demonstration. Our approach aims to incorporate task domain knowledge, where appropriate, by working in a space consistent with the constraints of a specific task. In addition, we also present an approach to using sensor feedback to learn a predictive model of the task outcome. This allows our system to learn the proprioceptive sensor feedback needed to monitor subsequent executions of the task online and abort execution in the event of predicted failure. We illustrate our approach using two example tasks executed with the PR2 dual-arm robot: a straight and accurate pool stroke and a box flipping task using two chopsticks as tools.

Author(s): Pastor, P. and Kalakrishnan, M. and Chitta, S. and Theodorou, E. and Schaal, S.
Book Title: IEEE International Conference on Robotics and Automation (ICRA)
Year: 2011
Department(s): Autonomous Motion
Research Project(s): Autonomous Robotic Manipulation
Associative Skill Memories
Bibtex Type: Conference Paper (inproceedings)
Address: Shanghai, China, May 9-13
Cross Ref: p10446
Note: clmc
URL: http://www-clmc.usc.edu/publications/P/pastor-ICRA2011.pdf

BibTex 

@inproceedings{Pastor_RAIIC_2011,
  title = {Skill learning and task outcome prediction for manipulation},
  author = {Pastor, P. and Kalakrishnan, M. and Chitta, S. and Theodorou, E. and Schaal, S.},
  booktitle = {IEEE International Conference on Robotics and Automation (ICRA)},
  address = {Shanghai, China, May 9-13},
  year = {2011},
  note = {clmc},
  doi = {},
  crossref = {p10446},
  url = {http://www-clmc.usc.edu/publications/P/pastor-ICRA2011.pdf}
}