One of the most important topics of Industry 4.0 is Predictive Maintenance, which helps companies reduce unexpected breakdowns in their equipment. The sophisticated instrumentation present in industrial machinery allows the collection of an enormous amount of data. The heterogeneous nature of this data makes it difficult to interpret, correlate, and use in a practical way in decision making. Additionally, many companies are missing data from system failures which are critical to establish robust prognostic algorithms.
In this seminar we will show tools to label data, design condition indicators and estimate the remaining useful life (RUL) of an electromechanical device. You can analyze and label machine data imported from local files, cloud storage, and distributed file systems. Creating a digital twin of your machine using multidomain physical networks is critical to simulate and generate synthetic datasets for fault conditions that would be too dangerous and costly to measure during a lab experiment. Using techniques and apps to select and calculate relevant features, a predictive model will be trained from these data to estimate the remaining useful life (RUL) of the device.
Finally, we'll show you how to deploy these algorithms in both enterprise environments (servers, cloud, etc.) and embedded devices using C/C++/IEC61131 code generation.