An Evaluation of GUI and Kinesthetic Teaching Methods for Constrained-Keyframe Skills

Abstract

Keyframe-based Learning from Demonstration has been shown to be an effective method for allowing end-users to teach robots skills. I propose a method for using multiple keyframe demonstrations to learn skills as sequences of positional constraints (c-keyframes) which can be planned between for skill execution. I also introduce an interactive GUI which can be used for displaying the learned c-keyframes to the teacher, for altering aspects of the skill after it has been taught, or for specifying a skill directly without providing kinesthetic demonstrations. I compare 3 methods of teaching c-keyframe skills: kinesthetic teaching, GUI teaching, and kinesthetic teaching followed by GUI editing of the learned skill (K-GUI teaching). Based on user evaluation, the K-GUI method of teaching is found to be the most preferred, and the GUI to be the least preferred. Kinesthetic teaching is also shown to result in more robust constraints than GUI teaching, and several use cases of K-GUI teaching are discussed to show how the GUI can be used to improve the results of kinesthetic teaching. The results indicate the benefit of K-GUI teaching over kinesthetic teaching, and suggest that more robust constraints may be needed to more flexibily represent skills and allow for a more intuitive GUI.