Processing Big Data in Streaming for Fault Prediction: An Industrial Application

In the current data-driven industrial scenario, Big Data processing plays a leading role in enhancing business performance. An ever increasing number of working machines are equipped with smart devices (such as sensors and actuators) which are in charge of monitoring machine status in real time and implement corrective actions before the workpiece quality is compromised or machine is damaged. However, many manufacturing companies do not take advantage of the use of Big Data coming from their production systems. In some cases, Big Data analytics is an un-explored issue since it is considered time and resources consuming. Moreover, the real benefits of processing, in real time, industrial data are usually underestimated. The European project TOREADOR wants to extend and facilitate the diffusion of Big Data analytics within industrial contexts, in order to generate greater value for companies. Focusing on the aerospace components manufacturing process (as one of the case studies of the Project), the paper aims to describe the main developments and lessons learned by the customization of TOREADOR Big Data analytics platform to process and analyze data from CNC machines sensors.