Machine Learning Pathway for Harnessing Knowledge and Data in Material Processing
- PDF / 1,081,320 Bytes
- 13 Pages / 595.276 x 790.866 pts Page_size
- 87 Downloads / 159 Views
Copyright 2020 American Foundry Society https://doi.org/10.1007/s40962-020-00506-2
Abstract Artificial intelligence (AI) is integral to Industry 4.0 and the evolution of smart factories. To realize this future, material processing industries are embarking on adopting AI technologies into their enterprise and plants; however, like all new technologies, there is always the potential for misuse or the false belief that the outcomes are reliable. The goal of this paper is to provide context for the application of machine learning to materials processing. The general landscapes of data science and materials processing are presented, using the foundry and the metal casting industry as an exemplar. The challenges that exist with typical foundry data are that the data are unbalanced, semi-supervised, heterogeneous, and limited in sample size.
Data science methods to address these issues are presented and discussed. The elements of a data science project are outlined and illustrated by a case study using sand cast foundry data. Finally, a prospective view of the application of data science to materials processing and the impact this will have in the field are given.
Introduction
to integrate connected factories within industry, decentralized and self-optimizing systems, and the digital supply chain in the information-driven cyber-physical environment of the fourth industrial revolution.1,2 The evolution toward Industry 4.0 is given in Figure 1.
The fourth industrial revolution that ushered the Internet of Things (IoT) and the Internet of Services (IoS) has come to be known as Industry 4.0. At the Hannover Messe in 2011, Germany launched a project called ‘‘Industrie 4.0’’ designed to fully digitize manufacturing. The larger vision of Industry 4.0 is the digital transformation of manufacturing, leveraging advanced technologies, and innovation accelerators in the convergence of IT (Information Technology) and OT (Operational Technology). The purpose is
International Journal of Metalcasting
Keywords: Industry 4.0, machine learning, smart factory, IoT, artificial intelligence, classification models, random forest, XGBoost, unbalanced, semi-supervised, dimension reduction, principal component analysis, feature importance, data standardization
The initial goals of Industry 4.0 typically have been automation, manufacturing process improvement, and productivity optimization. The more advanced goals are innovation and the transition to new business models and revenue sources using information technologies and
manufacturing processes are complex and appropriate data may not always be available for all parts. Accordingly, the need for advanced unsupervised or semi-supervised machine learning algorithms also exists.5 ‘‘The Landscape of Machine Learning’’ section describes these types of algorithms in detail. In this work, we want to show how these techniques can be used to answer the questions: How can we develop algorithms and apply AI/machine learning to processes where one does not have many defective, or otherwise labele
Data Loading...
