One focus here is to investigate the connection between process and pump response. The implementation stages include creating a sensor set-up for selected pumps in the Fraunhofer EMFT cleanroom so as to be able to log data at various points on the pumps and also set up a linked sensor node network complete with a secure Internet of Things (IoT) infrastructure.
Machine learning is to be deployed so as to detect irregularities in the sensor data. For this purpose, a data fusion of various sensor data is required in order to detect combinations and patterns, and software algorithms are needed to detect specific instability states.
Another aspect is encrypted wireless communication between the devices. A remote connection will make condition monitoring easier for operators in the factory. Information on the actual condition of a machine will be made available for retrieval. The Fraunhofer solution Industrial Data Space is to be used to prevent unauthorized access during data transmission – this is the new reference architecture for data storage in the area of networked industry automation (Industry 4.0). Sensor data from various units within the factory can be saved in a central database; access to this data is then limited by applying differing directives and access rights.