Building Predictive Maintenance Framework for Smart Environment Application Systems
Predictive maintenance is a powerful maintenance strategy that makes it possible to significantly reduce operation and maintenance costs of public, commercial and industrial environments. It is a complex data-driven process, which tries to forecast future states of company assets. On one hand it prerequisites condition monitoring of components on machine level. On the other hand it demands the integration of the collected data with other management information systems. Digitization and especially the advent of big data science bring along promising opportunities to create effective smart monitoring and predictive maintenance applications. The aim of this research is to examine the possibilities of a predictive maintenance framework based on the design principles of Industry 4.0 and recent developments in distributed computing, Big Data and Machine Learning. It introduces numerous enabling technologies such as industrial Internet of things, standardized communication protocols, as well as edge and cloud computing. Moreover, it takes a deeper look at data analytical techniques and tools, and analyses performance of well-known machine learning algorithms. Paper proposes architecture of a predictive maintenance framework based on existing software and hardware solutions. As a proof of concept, a real-life smart heating, ventilation, and air conditioning (HVAC) application system is created and tested to demonstrate the possibilities of the proposed PdM framework.