29. November 2019, 08:00 Uhr | Inka Krischke
Stefan Hornung: „In the long term, artificial intelligence will be an essential component of sensors for condition monitoring.“
One of the novelties at the Balluff booth is a sensor for condition monitoring. Stefan Hornung, Industrial Portfolio Manager at Balluff, explains what makes it so special.
Mr. Hornung, you advertise the ‚BCM‘ sensor as ‚globally unique‘ - on the basis of which properties?
Stefan Hornung: Due to the sensor‘s flexibility, it is ‚easy to select and use‘: The condition monitoring sensor is multifunctional and can be precisely adjusted to any application. Thanks to its compact housing, it is easy to mount anywhere - but above all, it offers the acquisition of four different measured variables, in just one device! It also uses IO-Link and offers data preprocessing and evaluation onboard directly in the sensor.
Which four variables does the sensor detect? And what do you mean by data preprocessing?
Stefan Hornung:Amongst others the sensor records the physical properties vibration over three axes, temperature, humidity and air pressure. The resulting data are already aggregated and pre-processed in the sensor. For example, it determines the RMS or effective values directly from the raw vibration data. The sensor also calculates mean values or peak-to-peak values. This eliminates the cumbersome and time-consuming transmission of raw data to separate evaluation devices or the calculation of useful values in external systems. Instead, meaningful information is available directly from the sensor, which can be used directly for analysis and evaluation.
Does the sensor select important and less important data before it passes them on?
Stefan Hornung: The parameterization of the sensor via IO-Link offers the user many possibilities: The content of the process data is freely definable. Up to five measured or pre-processed data can be selected and transferred cyclically. Depending on the application, the most important and meaningful data can be selected. Ultimately, the ‚BCM‘ makes an important contribution to Smart Data, since only the data that are really needed are communicated further. If further data and detailed information are required, these can be read out acyclically from the sensor at any time on request.
How is the sensor connected to the machine and plant environment?
Stefan Hornung: The Condition Monitoring Sensor uses IO-Link as an industrial interface, which greatly simplifies data transfer and parameterization because no special software or programming environment is required. Many users are already familiar with IO-Link and can therefore easily integrate the ‚BCM‘ into their machine and system environment.
Can you give us details on how the whole thing was technically implemented?
Stefan Hornung: It was a great challenge for our developers to integrate all functions and the individual sensors for recording the above-mentioned measured variables in such a compact housing. State-of-the-art microelectronics - so-called MEMS - play a major role here. These micro-electro-mechanical systems are tiny components that combine electronics, sensor technology and micromechanical structures on a single chip. A powerful microprocessor then processes the signals from these microsensors into useful data and information.
What do you think future sensors for condition monitoring will look like?
Stefan Hornung: Condition monitoring will continue to grow in importance in automation. This also means that the corresponding sensors will have to keep pace with this development. I therefore expect that the functionality and flexibility of sensors for condition monitoring will continue to grow. In the long term, artificial intelligence will be an essential component of such sensors: These sensors will then independently trigger the optimization of production and maintenance processes on the basis of the data they collect, thus further advancing the digitization of the modern factory.
And when can we expect Balluff‘s first sensor with AI functionality?
Stefan Hornung: Even though AI is currently a buzzword on everyone‘s lips, real AI functionality is above all a question of definition. Making a sensor more intelligent is one thing; defining the task and the sense for it is another. This is a topic Balluff is working on intensively; I don‘t want to reveal any more at this point...