The Future of Automation: From Field Device Through Modularization to Data Mining...

Related Vendors

DÜPERTHAL Sicherheitstechnik GmbH & Co. KG

TURCK GmbH

Elementar Analysensysteme GmbH

NAMUR e. V. c/o Bayer AG NAMUR-Geschäftsstelle

No Industry 4.0 Without Data Mining

Informations- and communication technology enable solutions today which would not have been possible or economical a few years ago. Data mining is an example of a technology the uses of which were known a decade ago, but were not implementable. In this connection Thorsten Pötter, Bayer AG, spoke about the challenges of an intelligent data analysis and offered a peep on the subject of the next Namur Annual General Meeting.

His thesis: “Information, not data, is the gold of the 21st century.” Because smart business decisions require correct information at the right time and at the right place. He provided some food for thought on where we stand today. Thus for instance, information systems already exist in the total production chain today, which can delivery any amount of information.

It is relevant to ask the question whether it is possible to continue using this data source and if yes, how? An idea is the automation of the thought process which is described under the umbrella of manufacturing intelligence.

Gallery

Gallery with 13 images

There could be a global batch cockpit for the multiplicity of data, with the help of which the different data heads could be tapped. Comparisons with the past (retrospect) an online monitoring (present) or a production planning (future) can be displayed in it. One should always have in view the purpose of such a data analysis and what one actually wants to achieve with the data analysis. And finally, the display of the information is decisive for its acceptance in practice.

Data mining can actually be a gold mine. Pötter gave the example of a project of GE Digital, in which data from more than 35,000 machines was collected and 100 million flight data was processed per annum. The combination of analytical methods for high data volumes and physical modeling led to early and more accurate error detection. This helped reduce operative disturbances significantly.

(ID:44395201)