Predictive maintenance is a technique to predict the future failure point of a machine component, so that the component can be replaced, based on a plan, just before it fails. Thus, equipment downtime is minimized, and the component lifetime is maximized.
On the basis of real-time data, all relevant parameters of the machines involved in the manufacturing process are acquired and evaluated for anomalies by means of stream analytics. In a subsequent machine learning process, specific fault patterns and the causes of a problem are detected in good time. This results in fewer rejects and maximum availability over the entire life cycle of the production line. The requirements on machine operating times vary depending on the specific branch of industry and its product cycles. While this could be up to 30 years in the aerospace sector, it is a matter of just a few months in the case of fast-moving goods, such as smartphones. With the aim of enabling an accurate assessment of the future performance of the machine or one of its components, intelligent predictive maintenance systems interconnect the largest possible amount of data from decentralized sources for the purpose of analysis.