Mr. Jamal, with all the Industry 4.0 euphoria, are we forgetting to include the discipline of testing technology?

Rahman Jamal: That is indeed the impression we get: When does the topic of 'testing' really come up in the context of Industry 4.0? The Industry 4.0 discussion is usually reduced to the automation of machine systems and fieldbuses. However, a digitalization strategy is simply incomplete without the inclusion of test and production test data.

Testing - the forgotten discipline?

Rahman Jamal: Testing has always been just an appendage, invisible, a cost factor! But you can't do without it: we are now increasingly hearing from test departments: "We need Industry 4.0-capable test solutions."

How is the testing industry reacting?

Rahman Jamal: We know what the IoT does to testing: it makes it more complex. Now we need to think about how IoT technologies can benefit testing. And here we can say that automated testing can benefit considerably from IoT technologies. Incidentally, this also applies to smart production, i.e. Industry 4.0! As a provider, we need to leverage this potential.

Which IoT platform technologies are you talking about?

Rahman Jamal: The technologies that characterize IoT platforms include the management of IoT endpoints and their networking capabilities. Access to data and data ingestion - i.e. the prioritization of data sources, the validation of files and their meaningful forwarding. The visualization and analysis of IoT data and the creation and management of IoT applications.

And to what extent can Industry 4.0 benefit from IoT technologies?

Rahman Jamal: Well, there are many similarities between smart production and automated testing: large amounts of data have to be generated for complex machine systems in order to implement predictive maintenance, for example. We also have this with automated testing: large amounts of data are generated here, and asset and system management is required. Data analytics is also required for predictive maintenance and condition monitoring. The only difference here is that we are dealing with automated test benches and therefore with test data.

In short, automated testing - as well as agile production - could benefit considerably from IoT platforms and IoT technologies in the areas of system management, data management, data analytics, application development and application management.

A software-defined approach for each individual automated test bench is necessary in order to be able to decide where which data should ultimately flow.

© National Instruments

Why don't you break these topics down to the world of testing?

Rahman Jamal: Take a look at the topic of system management: IoT devices are inherently well networked and managed. In practice, however, this hardly applies to test systems, even though they are increasingly distributed worldwide. Many older measuring devices and test benches are not networked. There is also a proliferation of interfaces such as GPIB, LXI and serial protocols. There is also often no overview of where and in what condition the scattered systems are located. It is therefore difficult to understand which driver and software version is running on them and which different hardware variants are being used.

The situation is similar when it comes to data management: you will find a multitude of data formats and sources - be it parametric measurements or raw analog and digital signal curves in the time and frequency domain, which are generated by the various interfaces such as GPIB, VXI, LXI, PXI or data acquisition cards. In addition, the data is often stored in silos and in different standards. As a result, data is not 'visible' across the organization and valuable insights into some phases of the product lifecycle are simply missing.

If we now look at data analytics, we see that commercially available business analytics software is not designed for complex and multidimensional test data. For example, it lacks the usual visualization functions for tests and measurements, such as combined graphs of analogue and digital signals, eye charts, Smith charts and constellation diagrams.

However, if you ensure that the analytics software is designed for test data, there are many advantages. For example, a wide variety of analyses can be carried out - from simple statistics to sophisticated algorithms from artificial intelligence or machine learning methods. Common tools such as Python, R and MATLAB can also be integrated into the workflow.

In the area of test software management, on the other hand, the trend is clearly moving away from purely centralized desktop applications towards more distributed thin clients or apps on mobile devices. In the test environment, however, there is a tendency towards proprietary, organically grown central solutions. A well-structured modular test software architecture with test management, test code, measurement IP, device drivers, hardware abstraction layers and so on would be an excellent basis for utilizing the many advantages of cloud computing, for example.

So what needs to happen in the testing world?

Rahman Jamal: The basis for each individual automated test bench is first of all a software-defined approach with a high degree of modularity. The next level is the different categories of data. On the one hand, the test benches generate a lot of data: Application, measurement and parametric data. On the other hand, the test bench itself is also on the test bench! In other words, I also have to look at the health status of my test asset in order to improve the utilization of the test equipment, monitor the condition of test components and detect a possible failure in advance. Now we have the different data and can now use the IoT platform capabilities for automated testing. In other words, I can use a wide range of services for automated testing, from simple messaging services to sophisticated deep learning and AI algorithms that optimize my workflow. It's no coincidence that people talk about the future being the so-called 'Internet of Services'.

And on this basis, we can decide what we want to move to the cloud, what we want to pass on to the ERP systems, which reports we want to generate, all the things that are important for business decisions.

Are we talking about beautiful visions or can users already put some of this into practice today?

Rahman Jamal: We recently introduced SystemLink, an application software for managing globally distributed test systems - that's the direction we're heading in! This software addresses precisely the points mentioned here: system and data management, data analytics and test software management. With a central, web-based interface, it ensures the management of devices, software and data for more efficient processes and greater productivity.