Skip to main contentSkip to page footer

 |  Blog

From OT to IT - Data Mining as Basis for Digital Business Models

"Data is the new gold" or "Data is the oil of the 21st century" are quotes heard more and more frequently in the course of Industry 4.0. This expresses the increasing importance of data. After all, data is the basic prerequisite for digital business models. However, they are not simply "there", but must first be collected. We show how data can be retrieved from operational technology (OT) and preprocessed before it is made available to the Cloud.

In our projects, the most common approach to successful data collection is through an edge gateway equipped with the appropriate software. With an edge software platform in place, the actual data collection is ready to go. Due to the large number of technologies, protocols and systems found in the field, a standardized approach is usually not possible. In addition, the approach varies depending on the type of data. Typically, one starts "at the bottom", at the level of sensors and actuators.

Often, you also have to deal with sensors that do not have a digital interface, but feed the measurement and control values into the automation system via physical variables such as currents and voltages. Such data can only be accessed via coupler components or a controller. 
If one or more of the numerous fieldbuses are used in the OT, the individual properties of the respective fieldbus technology must be considered. Practice has shown that almost every project requires an individual approach. 

The link is a controller (PLC) with an OPC UA server or other public interface where the data can be retrieved. Provided that the application running on the controller is designed to publish the data. This must be enabled and possibly programmed out by the PLC programmer. However, it is typical that a PLC performing control tasks cannot - or should not - be used for data collection. Often a separate system is used for data acquisition, virtually in parallel with the actual control system. In this way, unwanted feedback from data acquisition into the actual automation can be eliminated.

It may then be necessary to install additional hardware to provide access to the OT network. A (small) controller, preferably with an OPC UA server, can be used here. For IP-based fieldbuses, a standard computer without OT-specific interfaces may be sufficient.

Once the conditions of the OT have been clarified, they need to be "cast" in software. Software components that receive data from the OT are called OT connectors. They are implemented to integrate with the edge software platform in use. They retrieve data from an automation component (preferably via OPC UA) and make it available to the message bus that is part of the edge software platform. Other software components can then retrieve and process this data.

Before the data is forwarded to the Cloud or a neighboring system, it is usually pre-processed. For example, filter applications can omit irrelevant data. As a result, not all data is forwarded, but only selected data series and analysis results.

Our colleague Michael Heller shows you how to send this data to the Cloud and process it there in his white paper: Whitepaper: Data Mining "on the Edge" as a Basis for Digital Business Models. Our team of experts will be happy to advise you personally. 

About the author


Michael Heller is a computer scientist with a passion for automation and operational technology (OT). As a group leader and expert in the field of Industrial IoT and Edge, he is passionate about the "things" of Industrial IoT and their connection to neighboring systems, such as the Cloud, at M&M Software.

Created by