Programming of an Industrial Robot Using Demonstrations and Soft Computing Techniques

Abstract

In this work, trajectory planners are developed to program an industrial robot using human demonstrations aided by artificial neural network and human hand synergy approach. These planners are developed to acquire trajectory from human demonstrator and convert it to robot trajectory. The planner based on artificial neural network utilizes kinematic model of human and robot hand for mapping whereas human hand synergy approach is used by second planner for mapping. The proposed approach is implemented to an industrial robot in a designed scenario. The performance of both the planners is compared using an established metric. Experimentations and analysis of results revealed that both the approaches generalize well in the tested scenarios. Moreover, the planner adopting using human hand synergy for mapping shows better performance than kinematics based planner.