Learning from demonstration algorithm for cloth folding manipulator

Abstract

The aging population and high labor cost have accelerated the need for service robots for household activities. The household activity being diverse robot training over dynamic environments is required. Our work aims in developing an algorithm for teaching a manipulator to perform a task. Teaching pendant simplifies the process of teaching for the end user and also trains manipulator to a user specific problem. In this work, we present an approach to teaching a manipulator using Learning from Demonstration technique. Folding a cloth is taken as activity to teach the robot. Amrita Dual Anthropomorphic Manipulator (ADAM) is used to test the proposed algorithm.