Fast Data-based Production Support System

OData support
Supervisor:
Dr. Forstner Bertalan
Department of Automation and Applied Informatics

In the last couple of years, a new industrial revolution, called Industry 4.0 has started. Its main goals are to increase the level of automation, integrate different devices and networks and to analyze the vast amount of data to solve different problems. Besides this, there are a lot of different big data and machine learning technologies and methods being developed. The question is what sort of problems can be solved using these modern technologies in a production line environment. One of the main challenges is the maintenance of the different devices. In a modern production line environment, there are several components working together to produce the different products. The maintenance of these components is not trivial. The traditional approaches to maintenance can result in unexpected outages. One of the possible solutions to this problem could be predictive maintenance. During predictive maintenance, the main goal is to predict when the failures will occur using the data from different sensors and applying machine learning. Recently, deep learning techniques have become incredibly popular in a lot of different applications, e.g. face detection, voice recognition, object detection etc. In this paper, the design and implementation process of a system will be presented, which is able to solve the problem of predictive maintenance using modern big data and deep learning technologies.

Downloads

Please sign in to download the files of this thesis.