Predicting anomalies in automated production based on artificial intelligence methods

DOI №______

Abstract

This article is about exploring the use of Machine Learning and Deep Learning in industrial production. The issues of transition to Industry 4.0 and Maintenance 4.0 implementation are covered. As more industries move to Industry 4.0 and Maintenance 4.0, there is obviously a huge need for predicted maintenance of huge data sets from sensors installed on industrial equipment.
Machines learn to see and classify images, they can recognize the text and numbers in those images, as well as people and places. And computers do not just detect written words, but also take into account the context of their use, including the positive and negative shades of emotions.
A technology solution is required for businesses to adopt and scale predictive industrial IoT on their facilities. The Auto-MDL system was developed for this purpose — a non-standard solution for industrial enterprises moving to Industry 4.0, using a similar level of resources and capabilities when operating under old conditions.
The scientific problems of this problem were determined and the current state of use of methods of solving the forecasting problem was investigated. The problems of using neural networks were investigated in more detail and clustering algorithms were investigated.
The article describes the use of machine learning methods in anomaly prediction tasks, Industry 4.0 and Maintenance 4.0. Machine and depth learning algorithms were investigated and the results of implementation of machine learning algorithms to detect anomalies in real industrial production were described.
The development of automated software architecture based on machine learning algorithms for predicted maintenance is described. An industrial real-world example using an automated manufacturing system was described.

Keywords: Machine Learning; Deep Learning; Artificial Intelligence; IoT; Internet of Things; Industry 4.0; Maintenance 4.0.

References
1. Хель І. Індустрія 4.0: що таке четверта промислова революція? // Hi-News, 2015.
2. Machine Learning with AWS. Pack Publishing: November 2018.
3. Smola A., Vishwanathan S. V. N. Introduction to Machine Learning. United Kingdom: Cambridge University Press, 2008. 226 р.
4. Hozdić E. Smart factory for industry 4.0. Slovenia: University of Ljubljana, 2015. 35 р.
5. Google Developers, Machine Learning. Google Engineering Education Team.
6. Presser M., Institute A., Höller J. Inspiring of Internet of Things. Ericsson, Raaskot, Denmark, 2011. 44 р.