Application of Machine Learning for Failure Prediction in Manufacturing Process  
Author Ivan Zajacko


Co-Author(s) Ivan Kuric; Tomas Gal


Abstract This paper deals with research and development of an autonomous system for multicriterial predictive diagnostics which allows us to respond quickly enough to emergence of defects in the production process. It describes the diagnostic methods used in the designing of a Machine Learning system that uses neural network to process data from the production line. We focus on analyzing the vibrations in the production line that adversely affect the quality of the final product and development of a model that can predict the emergence of these vibrations and their mitigation by proactively controlling parameters of the production line. Our diagnostic and prediction model was integrated into the industrial control system of the production line, and we evaluated the performance of the developed control code.


Keywords predictive diagnostics, machine learning, neural networks, vibration analysis
    Article #:  DSIS19-46
Proceedings of ISSAT International Conference on Data Science & Intelligent Systems
August 1-3, 2019 - Las Vegas, NV, U.S.A.