Artificial Intelligence Platform Proposal for Paint Structure Quality Prediction within the Industry 4.0 Concept

Abstract This article provides an artificial intelligence platform proposal for paint structure quality prediction using Big Data analytics methodologies. The whole proposal fits into the current trends that are outlined in the Industry 4.0 concept. The painting process is very complex, producing huge volumes of data, but the main problem is that the data comes from different data sources, often heterogeneous, and it is necessary to propose a way to collect and integrate them into a common repository. The motivation for this work were the industry requirements to solve specific problems that cannot be solved by standard methods but require a sophisticated and holistic approach. It is the application of artificial intelligence that suggests a solution that is not otherwise visible, and the use of standard methods would not give any satisfactory results. The result is the design of an artificial intelligence platform that has been deployed in a real manufacturing process, and the initial results confirm the correctness and validity of this step. We also present a data collection and integration architecture, which is an integral part of every big data analytics solution, and a principal component analysis that was used to reduce the dimensionality of the large number of production process data.