In this chapter we present the realization of a prototype infrastructure aiming at providing a useful framework to collect and elaborate information in a big data and IoT environment. The work presents a novel approach related to predictive maintenance for automatic packaging machines, dealing with condition monitoring of mechanical components. The knowledge of the state of machinery parts is crucial to trigger dynamic scheduling of their servicing before they are worn out or get corrupted, saving time and money. In this fashion, condition monitoring, also known as incipient fault diagnosis, has a key role in the estimation of components’ condition and their remaining working time.
This project is within the Industry 4.0 framework. It refers to the need to perform predictive maintenance on machinery working with heavy-duty cycles. This is a major request from machinery vendors and their customers: let the machine inform the user (and its seller) about its components state. Diagnosis of faulty or corrupted components have become complex due to growing machinery complexity. In this fashion, the University of Bologna and its industrial partners are developing predictive maintenance and diagnosis techniques by exploiting the Internet of Things and big data concepts of Industry 4.0.
Rules and models are needed to achieve effective incipient fault diagnosis especially if they ...