TwinCAT Analytics

TwinCAT Analytics is a tool for data analysis from one local or from many distributed machines. The TwinCAT Analytics product family represents the entire workflow of such an analysis and includes a wide variety of algorithms for the analysis of live data as well as historical data. The data that is processed into valuable information and forms the basis for decision-making with the help of TwinCAT Analytics represents the "raw material" for this workflow.

In the following sections, we will take you along the path of data – from the collection of raw data to possible measures and optimizations – of course completely based on TwinCAT Analytics.

Step 1: Raw data acquisition

Where does the data actually come from?

The quality of an analysis correlates with the quality of the data ("garbage in – garbage out"). Therefore, raw data collection represents the starting point of the data value creation process. For successful analysis, it is relevant to recognize which data is needed for which purpose and how fast changes have to be recorded (sampling rate). TwinCAT Analytics offers different options for the continuous acquisition of the process image. With the help of high sampling rates, large amounts of data can be acquired in real time – for high data quality as the basis for successful analysis and model building. The use of the TwinCAT 3 Analytics Logger is particularly effective, but other well-known and proven technologies and products, such as the TwinCAT 3 Scope View or the TwinCAT 3 Database Server, can also be used.

Step 2: Data storage and communication

How does the data get from the collection location into my analysis?
Where is the data stored and what are the access options?

Following data collection or directly during the collection of data, it must be transferred or stored. Our products offer you a wide range of possibilities. Live data can be transmitted directly via ADS or MQTT for live analysis. If the data was initially historicized on the local target system or in the network with the TwinCAT 3 Analytics Storage Provider, analytics files can be transferred to another system and archived at a later time. Import to databases or export to other file formats are also possible.

Step 3: Data preparation and exploratory (pre-)analysis

How is the data processed and made available for the analyses?
How is a problem-specific analysis designed?

In many cases, the path to a successful analysis leads first through data pre-processing and exploratory pre-analysis of the data. This is because it is not always clear from the outset which data streams need to be processed using which algorithms in order to arrive at a meaningful analysis chain that delivers the desired results and recommendations for action. Therefore, relevant data streams and, if necessary, time periods and events must first be identified. The Data Scout, which is integrated in the TwinCAT 3 Analytics Workbench and in the TwinCAT 3 Analytics Service Tool, can help you with this. This provides a quick overview of individual data collections and also allows data from different data collections to be cut and linked to form a new data stream. Once the appropriate data has been selected, suitable algorithms and links must also be found. For this purpose, the TwinCAT 3 Analytics Workbench with the integrated configurator offers the best possibilities. You can easily drag and drop the various algorithms from the toolbox into the configurator, link them with your data and other algorithms, and parameterize them according to your ideas. If you need user-specific algorithms, it is also possible to extend the existing algorithms with the new Analytics Lambda functions, which can be implemented with the help of a wizard.

Step 4: 24/7 data collection, analysis and integration into operations

In which way can the created analysis be used permanently and continuously?

Following the identification of relevant data, you can select it for continuous data collection. For this purpose, the TwinCAT Analytics Logger can be configured again; it writes the selected data streams during the runtime of the control and thus enables continuous data acquisition. Automatic code generation allows you to generate PLC code from the problem-specific analysis chain previously created in the Workbench. This can then also be permanently integrated into the operation and perform the desired analysis of the recorded data 24/7. This makes it possible to detect possible error cases promptly and with foresight, and to uncover correlations as well as optimization potential even over long periods of time.

Step 5: Visualization of the analysis results

How can the analysis results be clearly presented?

The next logical step with regard to a permanent integration of the analysis into the operation is to continuously visualize the machine condition as well as the analysis results. The One-Click Dashboard makes this a very short step and a particularly easy one for you as a user to implement. You can make configuration settings for the later dashboard even during the design of the analysis in the workbench. The dashboard is then automatically generated and can be flexibly customized. This gives you an interpretable visualization of your data and analysis results integrated into the TwinCAT HMI, which is also an intuitive user interface.

Step 6: Measures and optimizations

How can measures be derived and optimization potential be exploited?

Continuous analyses can be used to derive ongoing measures that shorten machine downtimes and extend the service life of machines and systems, as well as increasing product quality. In addition, optimization potential, e.g. for optimizing processes or machines, can be derived from the "on demand" analyses in the TwinCAT 3 Analytics Workbench or the TwinCAT 3 Analytics Service Tool and integrated into production. When actions and consequences are derived at the variable level, they can also be written directly back to the control, closing the analysis-action loop.

Learn in the following video how TwinCAT Analytics products cover all the individual process steps of data value creation. All the above steps, from raw data acquisition, data storage, data preparation, 24/7 data recording, to visualization and optimization at variable level, are implemented by TwinCAT Analytics software components. And all this without much engineering effort or sophisticated PLC code programming. Most application scenarios can be mapped by already implemented algorithms and supporting wizards.

Benefit from the simplicity of the technological implementation of data analysis with TwinCAT Analytics and use your data profitably with the development of new business models for your company which is made possible as a result.

You will find useful instructions and tips on how to work with selected TwinCAT Analytics software functions in these short technical videos.