Meta-learning for enhancing machines' intelligence in manufacturing

Machine learning is usually considered a branch of artificial intelligence that enables machines (computers) to learn from data and improve their performance without explicit programming. The paper presents a review of research on meta-learning, also called double and semi-double-loop machine learning, based manufacturing. The paper provides a selected list of meta-learning definitions and generic models and discusses implementation architectures. Following, a set of meta-learning applications in manufacturing is presented, addressing three sub-domains: the basic "classical" processes in manufacturing systems, the manufacturing systems, including some advanced concepts such as smart manufacturing, advanced communication, and Industry 4.0, and manufacturing eco-system as a generic name for a number of application sub-domains that are directly or indirectly related to manufacturing and that could provide virtually easy knowledge transfer to manufacturing. In the last part of the paper, some future research and development directions are given.