The implementation of the Industry 4.0 idea is progressing very slowly. For eight years now, Industry 4.0 has been driven through the streets. Every self-respecting player has presented ideas, concepts and products for or around the hype: Cloud solutions, servers, node and edge devices through to Industry 4.0 sensors.

Fueled by the large number of 'connected' devices that the market is expected to produce and the huge amounts of data that need to be processed, many providers have seen the dollar signs literally pop up in their eyes. - No one wanted to be late to the market with corresponding products.

And so each provider virtually disappeared into its development department with an initial idea and came up with what it considered to be the best solution. Or, on the basis of their own customer base and supposed market power, they cooked up their own little soup to secure their traditional part of the business. According to the motto: from the hardware to the communication protocol to the cloud - my solution is the best!

The result is that with the wide range of offerings at all levels that the Industry 4.0 wave has now created, the customer has one main problem: deciding what to choose!

The large numbers problem

In general, a large-numbers situation is positive for the customer. A wide range of products with many competitors enables them to negotiate down the price. However, this approach does not take place in the market for Industry 4.0 products, or only to a very limited extent. Why is this the case? The customer is indeed faced with many offers, but these are not 1:1 interchangeable. In this case, there is an atypical large-numbers problem. The customer has the problem of deciding on an offer within this range of offers due to the wide but not directly interchangeable offer.

Where does the problem come from? The offers at the various implementation levels are not directly complementary due to a lack of standards, i.e. they are not in direct competition due to the lack of interchangeability. This is because the providers are currently still trying to secure the theoretically maximum market share by differentiating themselves as much as possible. The problem here essentially lies in the communication technology between the individual implementation levels: While the path from the sensor to the edge device is still well defined and standardized with the established serial interfaces or USB, the situation is different at the levels above.

How does the data get from the edge to the node? And further: How does the data get to the server and then to the cloud? Or does the data go directly to the cloud? And if so, which one? The combination of these questions paired with the respective broad range of options results in a complex decision matrix.

What's more, the customer is entering uncharted territory at almost all levels of an Industry 4.0 implementation. Mechanical engineering, which is considered conservative, often lacks the courage to take a 'first position'. The risks associated with the complex implementation of Industry 4.0 often seem too great compared to the potential positive effects. And an orderly partial entry is ruled out due to a lack of future security - due to a lack of standardization. Even various funded lighthouse projects such as the 'Cluster of Excellence Integrative Production Technology for High-Wage Countries' or 'Autonomics' have not yet eased this fear barrier due to their low profile and focus on large-scale projects.

Standard = openness

Openness - yes, it has been and continues to be signaled in the automation industry. In the past, however, it was non-existent on closer, critical inspection. To date, there has simply been a lack of real openness in the form of standards!

Yet the industrial PC sector - with standards such as VME, CompactPCI, ETX or Smarc - proves that the direct route via a standard is simpler and faster. Even if this path is often not easy. In the case of ETX, for example, it started with the search for partners, as ETX was driven by the manufacturer Jumptech at the time and for a long time had a 'proprietary' feel to it, but was ultimately adopted by more and more market players and finally accepted on the market as a 'quasi-standard'.

And perhaps the automation industry still has to learn that a real standard cannot be created by sheer size. Even companies such as Siemens or Deutsche Telekom are finding this difficult. However, it would also be sad if standards were always created by vendor power, as the best solutions would hardly or not at all be heard on the market due to the lack of size of the innovator and would disappear from the scene without a sound.

On the customer side, openness in the form of standards is therefore required. The demand for at least one 'second source' can hardly be realized in the current constellation across all levels of Industry 4.0. And very few people are prepared to go down the rocky road of putting together their own second source in a modular system due to the complexity involved. Replacing a standardized Smarc module on the carrier developed for the application is not enough. With Industry 4.0, there are simply a lot of interlocking gears, which makes such a project so complex.

Another important point is the quick and easy integration of the data into a cloud. But which cloud to use? And to what extent are the cloud offerings comparable in terms of connection and handling? At first glance, it seems obvious to fall back on the offerings of the top dogs Microsoft (Azure) and Amazon (AWS - Amazon Web Services). However, questions such as "Does my data belong to me?" and general security concerns are the brakes here. In addition, it is not only the production disciplines that decide on these IoT implementations, but also the company's IT department needs to be brought on board.