Mr. Jamal, you describe the Industrial Internet of Things - IIoT for short - and the associated machine learning as 'the' big trend in industry. Where do we currently stand in this respect?

Jamal: To quote one of the latest analyses - from Accenture - 95% of corporate decision-makers expect their company to jump on the IIoT bandwagon within the next three years. Unfortunately, in the general trend hype, the upcoming challenges in practice usually fall by the wayside.

What are they?

Jamal: IIoT applications are constantly changing. This means that security updates need to be installed constantly. It is also important to keep IIoT networks up to date. In abstract terms, two worlds collide, namely IT and OT, i.e. operational technologies, with completely new challenges for providers and users alike.

The growing number of OT systems brings with it veritable floods of data - what we at National Instruments call 'Big Analog Data'. It is undisputed that this data contains important information about machine failures and production errors. However, the amount of data is growing so quickly that it can no longer be handled with today's expertise and conventional tools. This is where machine learning comes into play - an area of AI that, simply put, is all about the 'artificial' generation of knowledge from experience.

Regarding the huge amounts of data - what is still missing today to cope with them?

Jamal: Quite simply, the right tools for topics such as remote administration, software configuration and data management. We are now tackling these tasks by addressing these challenges with tools such as SystemLink, the Data Management Software Suite and the Labview Analytics and Machine Learning Toolkit.

What is behind these tools?

Jamal: The 'Data Management Software Suite' is a complete solution for company-wide data management. It contains server-based software that is designed to quickly locate required data in a network. This is achieved through automatic indexing. Also included is client software specially designed to meet the needs of scientists and engineers for quickly finding, viewing, analyzing and evaluating measurement data without the need for additional tools. Finally, the third component provides the user with server-based software for the efficient handling, analysis and processing of large volumes of data. The Data Management Software Suite therefore comprises the most important software tools from NI for handling large amounts of data.

What is the Labview Analytics and Machine Learning Toolkit?

Jamal: With the Labview Analytics and Machine Learning Toolkit, machine learning can be implemented directly at the edge of condition monitoring applications for predictive maintenance. The toolkit enables the integration of predictive analytics and machine learning in Labview. It contains modules that allow the training of machine learning models that can be used to recognize patterns in large amounts of data. This is done by recognizing anomalies, classifying and clustering algorithms. The models can then be used for pattern recognition in new data on Windows computers or target systems with NI Linux Real-Time.