Industry 4.0 – Much Ado about Nothing?
We were very pleased to be able to welcome Dr. Missbach as speaker and guest on the topic Industry 4.0 at our 7 IT Forum in Igls.
Dr. Michael Missbach is known for his many books on SAP infrastructures and the Cloud. Since this year, Mr. Missbach has been working for Freudenberg-IT where, besides HANA, he is looking after topics such as Industry 4.0 and big data.
Because of his interesting talk on the forum titled „Internet of Things for the manufacturing industry? – practical examples and problem solving” we are looking forward to an exciting subsequent interview on the subject of Industry 4.0 with a quite critical perspective.
What is exciting about the topic Industry 4.0?
It is quite a paradox fact of the actual situation that many media permanently publish articles that predict the downfall of at least certain branches of the industry if not of the entire western civilization in case not all of us will digitalize everything as fast as possible. Here it is postulated that you only have to accumulate huge amounts of data in order to miraculously get to new findings.
In reality, you have to find such projects especially at medium-sized companies with a magnifying glass, although these companies would actually be predestined for it because of their concentrated technical expertise.
It is interesting to notice for me as a qualified mechanical engineer that it is pretended in all stories that until recently neither sensors nor the utilization of measured values for the optimization of equipment and processes were in existence. This even though James Watt already installed into his machines a centrifugal governor, which he invented in 1788, as well as an accurate sensor for the rotation speed and on top of that robust and fully automated controls. From this perspective, it does not come as a surprise that medium-sized companies regard the doomsday scenarios as well as the salvation that was promised by IT quite skeptically.
What is in fact new at Industry 4.0?
Largely unnoticed a lot of things actually changed over the last couple of years fundamentally. Thanks to mobile telephones, high-grade sensor technologies and worldwide connectivity have become available at relatively low costs. Today you will find more sensors in each smartphone than previously in a whole machine park. Freudenberg sealing technology developed for example the “Simmerring” that enables finding out the exact position of the crankshaft via a helical magnet field that is embossed on the rubber lip. On top of that, we have new data sources such as the analysis of product reviews by customers on the sites of online traders.
Even old machinery with “mechanical” displays, like for example the water level in a boiler, can be “digitalized” fairly cheaply without interfering with the machine by placing a webcam in front of the display and by using the web service of FIT, which will “read” the desired measurement value from the video stream.
Are there application examples that are proven already?
If you look closely enough then yes – but they are not very spectacular most of the time, like a smart city sponsored by tax money. Often you will find them at unexpected locations.
A fantastic example of an application no one ever thought of before is the Sonolyzer app of KSB. This app “listens” if a pump with asynchronous AC motor is running in the optimal load range. All you have to do is to place your smartphone with the activated app to the fan.
A further example is the open pit mining operation that evaluates the progression of torque at the rear axles of the heavy-duty trucks. The more irregular the progression of torque, the more likely that the gravel road has become a washboard track. On time grading of the roads will not only lead to significant savings in fuel, but will also increase the lifetime of tires, drive shaft, gear box and engine as well as giving a bigger hauling volume due to the higher driving speed.
What has Industry 4.0 to do with big data?
Here too the differentiated view of a practical expert is needed. Not every mid-sized company operates a hadron collider and thus is generates Petabytes of data. To analyze the above-mentioned product evaluations of customers you do not need to keep data on your own system for instance as the online traders already saved them.
Even when a production line for tiny medicine bottles, that is churning out 1 bottle per second, is capturing 150 data points per second, then – at 1.5 to 2KB per dataset – this amounts only to 52 GB per year even when it is 24 hours in operation.
During the analysis of railway tracks underneath a moving train with 35,000 datasets per hour, the measurements of more than 100 days will find space even on a MicroSD card with 16GB. To suit mid-sized companies, Freudenberg-IT is therefore offering the use of its big data platform in small “slices” of 1TB each.
But quite different to what the articles in the media make you believe, it is not the purpose of the big data to enlarge the “haystack” so that you can hope to find more needles in it. It is more a matter of suitable tools for analysis and the experts that can handle them. Hence, at the topics Industry 4.0 and big data that one has an advantage, who paid attention to the statistics lecture during his study and who recognizes the importance of the Gaussian normal distribution curve and the correlation coefficient.