Development of adapting gradient method for contour tracking in industrial robot application

Abstract

This paper describes the development of adapting gradient algorithm in order to automate the manual programming process. The outcome of this work is to make industrial robot capable of doing self learning to track contour for industrial application with minimal human involvement. Conventional industrial robot becomes a smarter machine which can exhibit intelligent behaviors such as obstacle avoidances, contour mapping, i.e. responds to sensory signals in accordance to changing environmental conditions and requires very simple programming. This new concept has been developed and tested on real industrial robot.