Robot Manipulation with a Human in the Loop - Trajectory Recording and Playback

Abstract

This project concerns the development of a system which allows tasks to be taught to a robotic manipulator using learning by demonstration. A subsystem calculates the accuracy required of the manipulator and determines which portion of the learned task it is able to perform. Therefore, the system can avoid performing tasks that require accuracy beyond the capabilities of the manipulator. The first sections of this paper are written by the entire project group and discuss the market and IP strategy suggested if the idea were commercialized. The final sections are written by myself and describe the trajectory recording and playback components of the project in technical detail.

Cite this paper

@inproceedings{Schweigert2015RobotMW, title={Robot Manipulation with a Human in the Loop - Trajectory Recording and Playback}, author={Sebastian Schweigert and Jiewen Sun and James Su and Sunil Srinivasan and Mark Jouppi and SEBASTIAN SCHWEIGERT}, year={2015} }