Efficient data analysis can also be applied in many other areas of production that involve complex technical systems. In fact, it can be used wherever sensors supply suitable data. Digitization leads to a big increase in the number of installed sensors, and Internet of Things applications are also being used to a growing extent. “That’s why Data Science with a focus on machine learning has extremely high economic potential,” says Warnke.
This applies to mechanical engineering in particular, because a machine that unexpectedly stops working generates unnecessary costs. However, the machines also supply huge amounts of information about why they stop working. How can this data be used to shorten maintenance times and make production more efficient?
This question is the focus of the current development of Uptime by the Data Science specialists at Körber Digital. Answers to this question are provided by the “classification” of the data using artificial intelligence (AI) technology. This system also forms the basis for cow diagnoses, because it analyses whether the values for milk temperature and milk amounts indicate that a cow is ill. The AI aims to detect patterns in a manner similar to how a spam filter determines whether an e-mail from an unknown sender is spam or not on the basis of the words it contains. With regard to machine downtimes, this means that if there is sufficient data that has been analyzed by AI, it can recognize the fault on the basis of the data scenario and identify the cause of the problem as early as the moment when the machine shuts down.