Alexander Skoglund, Tom Duckett, Boyko Iliev, Achim J. Lilienthal and Rainer Palm
Teaching by Demonstration of Robotic Manipulators in Non-Stationary Environments
Proc. IEEE Int. Conf. on Robotics and Automation (ICRA), 2006, pp. 4339 - 4341
Abstract: In this paper we propose a system consisting of a manipulator equipped with range sensors, that is instructed to follow a trajectory demonstrated by a human teacher wearing a motion capturing device. During the demonstration a three dimensional occupancy grid of the environment is built using the range sensor information and the trajectory. The demonstration is followed by an exploration phase, where the robot undergoes self-improvement of the task, during which the occupancy grid is used to avoid collisions. In parallel a reinforcement learning (RL) agent, biased by the demonstration, learns a point-to-point task policy. When changes occur in the workspace, both the occupancy grid and the learned policy will be updated online by the system.
Download:
Paper: [PDF (0.9MB)]
@INPROCEEDINGS{Skoglund_etal:ICRA:2006,
  AUTHOR = {Skoglund, Alexander and Duckett, Tom and Iliev, Boyko and Lilienthal, Achim~J. and Palm, Rainer},
  TITLE = {Teaching by Demonstration of Robotic Manipulators in Non-Stationary Environments},
  BOOKTITLE = {Proceedings of the IEEE International Conference on Robotics and Automation (ICRA)},
  YEAR = {2006},
  MONTH = {May 15 -- 19},
  ADRESS = {Orlando, Florida},
  PAGES = {4339--4341}
}