Possibilities of Using Deep Convolutional Neural Network and Autoencoder Neural Network for Predictive Maintenance Tasks

Entering the era of Industry 4.0, artificial intelligence has also found a place in the manufacturing industry. In the last decade, there has been an increased interest in the topic of AI in manufacturing. The issues of maintenance and technical diagnostics are significantly linked to the issues of manufacturing and represent a significant area of interest for science and research. Maintenance has so far been associated with modern technologies and analytical methods. Greater emphasis is being placed on predictive maintenance models that use operational data on the current state of machinery and equipment, allowing potential problems in systems to be identified before failure or malfunction occurs. This issue is one of the basic parts of any production process and is an indication of the malfunctioning and faulty states of production systems. For these reasons, the present research deals with the analysis of the possibilities of using deep convolutional neural networks and autoencoder networks for the tasks of signal processing, prediction of operating parameters, and anomaly detection in hydraulic circuits.