Real-World Robotics: Learning to Plan for Robust Execution

Abstract

In executing classical plans in the real world, small discrepancies between a planner's internal representations and the real world are unavoidable. These can conspire to cause real-world failures even though the planner is sound and, therefore, proves that a sequence of actions achieves the goal. Permissive planning, a machine learning extension to classical planning, is one response to this difficulty. This paper describes the permissive planning approach and presents GRASPER, a permissive planning robotic system that learns to robustly pick up novel objects.

Cite

Text

Bennett and DeJong. "Real-World Robotics: Learning to Plan for Robust Execution." Machine Learning, 1996. doi:10.1023/A:1018274104280

Markdown

[Bennett and DeJong. "Real-World Robotics: Learning to Plan for Robust Execution." Machine Learning, 1996.](https://mlanthology.org/mlj/1996/bennett1996mlj-realworld/) doi:10.1023/A:1018274104280

BibTeX

@article{bennett1996mlj-realworld,
  title     = {{Real-World Robotics: Learning to Plan for Robust Execution}},
  author    = {Bennett, Scott W. and DeJong, Gerald},
  journal   = {Machine Learning},
  year      = {1996},
  pages     = {121-161},
  doi       = {10.1023/A:1018274104280},
  volume    = {23},
  url       = {https://mlanthology.org/mlj/1996/bennett1996mlj-realworld/}
}