by F Stulp, W Koska, A Maldonado and M Beetz
Abstract:
One of the most notable and recognizable features of robot motion is the abrupt transitions between actions in action sequences. In contrast, humans and animals perform sequences of actions efficiently, and with seamless transitions between subsequent actions. This smoothness is not a goal in itself, but a side-effect of the evolutionary optimization of other performance measures. In this paper, we argue that such jagged motion is an inevitable consequence of the way human designers and planners reason about abstract actions. We then present subgoal refinement, a procedure that optimizes action sequences. Subgoal refinement determines action parameters that are not relevant to why the action was selected, and optimizes these parameters with respect to expected execution performance. This performance is computed using action models, which are learned from observed experience. We integrate subgoal refinement in an existing planning system, and demonstrate how requiring optimal performance causes smooth motion in three robotic domains.
Reference:
Seamless Execution of Action Sequences (F Stulp, W Koska, A Maldonado and M Beetz), In Proceedings of the IEEE International Conference on Robotics and Automation (ICRA), 2007.
Bibtex Entry:
@inproceedings{stulp_seamless_2007,
author = {F Stulp and W Koska and A Maldonado and M Beetz},
title = {Seamless Execution of Action Sequences},
booktitle = {Proceedings of the {IEEE} International Conference on Robotics and
Automation ({ICRA)}},
year = {2007},
pages = {3687--3692},
abstract = {One of the most notable and recognizable features of robot motion
is the abrupt transitions between actions in action sequences. In
contrast, humans and animals perform sequences of actions efficiently,
and with seamless transitions between subsequent actions. This smoothness
is not a goal in itself, but a side-effect of the evolutionary optimization
of other performance measures. In this paper, we argue that such
jagged motion is an inevitable consequence of the way human designers
and planners reason about abstract actions. We then present subgoal
refinement, a procedure that optimizes action sequences. Subgoal
refinement determines action parameters that are not relevant to
why the action was selected, and optimizes these parameters with
respect to expected execution performance. This performance is computed
using action models, which are learned from observed experience.
We integrate subgoal refinement in an existing planning system, and
demonstrate how requiring optimal performance causes smooth motion
in three robotic domains.},
}