Industrial transfer learning

From science to practical applications

Abstract

Despite the considerable potential of machine learning for solving common problems in automation technology, there are few examples of its application in practice. In order to examine the reasons for this, we looked at four practical cases, and present possible solutions involving industrial transfer learning. However, various preconditions must be met for the largescale use of such approaches. The article concludes with a discussion of these prerequisites and offers suggestions about how they could be fulfilled.