Implementation of Machine Learning Method for Positioning Accuracy Improvement in Industrial Robot

Abstract

Implementing of modern artificial intellect (AI) and machine learning (ML) for existing machinery can add value to their existing capabilities and technical characteristics. Machine learning is next step towards new innovations and stronger competitions in the market. Implementation of (ML) in the area of robotics requires some analysis of existing methods in order of correct of implemented method. This article sums up machine learning methods used in industry and presents successful implementation of deep Q-learning algorithm, implemented in robot static accuracy improvement using variable carrying load. Improvement reaches 0.07 mm for initial value equal to 0.1 mm. Finally, conclusions on implementing ML methods are drawn.