Concept of a learning knowledge-based system for programming industrial robots
Alejandro Magaña, Philipp R. Bauer, Gunther Reinhart · Procedia CIRP · 2019
A major challenge for the use of industrial robots has been its programming, which requires expert knowledge. Offline programming tools, based on simulation models, are normally used to facilitate the robot programming. However, in many industrial applications the intervention of an operator is still necessary to correct the robot programs. This paper presents a concept based on a knowledge-based system (KBS) that integrates the capability to learn from manual adjustments conducted by an operator. Based on the learned data the KBS will be able to imitate the adjustments of the operator, enabling a full automation of the robot programming.