The Composable Software background section deserves some kind of conclusion. Two and a half years after its launch, some of the fog has lifted for me.
What the internet, cloud and AI have long used as a basic technology is now also available in the industry: container apps. The basis is real-time Linux, which has been available for a good ten years.
This could mean the beginning of the end for stolid, proprietary, monolithic software that is far too inefficient and no longer up to date. But it is not clear whether the industry will take this step quickly enough. At the moment, it is mostly in the process of benchmarking for years, as in the past. And in China, they are already on the move.
In reference to Gartner and their vision of a composable enterprise, I called the software that is now also possible and necessary in the industry “composable software”. This is because the containerized apps not only run on any open runtime, they – and their data – can also be used with each other without cumbersome conversion and without regard to the programming environment from which they originate. But above all: without regard for and without dependence on the hardware and its manufacturer. Composable, in other words.
One of the comparison tables from the Smart Automation market overview, which shows the openness of the platforms for collaboration.
Probably the most important development in recent years in this context is the emergence of open, Linux-based platforms for automation. They were not even mentioned in my introduction to “composable software”.
The manufacturers are, on the one hand, established automation providers who have recognized the limits of their previously closed and hardware-dependent OT and, on the other hand, start-ups founded specifically for the open platform business. The 13 providers are currently represented on the Smart Automation market overview created on this portal in 2024. They are – in alphabetical order – as follows:
Bosch Rexroth with ctrlX AUTOMATION, FLECS Technologies with FLECS, German Edge Cloud with ONCITE DPS, Hilscher Gesellschaft für Systemautomation with netFIELD, KEB Automation KG with NOA, KEBA AG with Kemro X, Lenze with Lenze NUPANO, Phoenix Contact with PLCnext Technology, SALZ Automation with SALZ Controller, TTTech Digital Solutions with Ubique, TTTech Industrial Automation AG with Nerve, WAGO with WAGO OS and WAGO ctrlX OS, and Weidmüller with u-OS and easyConnect.
All of them have their headquarters in Germany or Austria. As far as I know, there is currently no direct competition from either the USA or China. The local providers probably have a head start of around five years. At most. I wouldn’t be surprised if something similar were to come onto the market from China in the near future. Because the big competition for the industry comes from China, not the USA.
As a reminder, when Xi Jinping announced the launch of the first of three ten-year plans with “Made in China 2025” in 2015, he also made it clear that the Chinese strategy was based on the German Industry 4.0 initiative, not the re-industrialization efforts in the USA.
Open platform – is that possible?
In contrast to what was previously understood as a platform in the world of IT and industrial software, a completely different world is being established here. This time, platform really means an open basis for any application and its data exchange and use. Not for the targeted lock-in of customers, who should be made as completely dependent on a platform as possible. And above all, not for hardware lock-in.
Interestingly, this creates a win-win situation that would have been unthinkable with the traditional type of OT and IT. This is because openness increases competition by making providers more transparent. But at the same time, it allows a veritable tsunami of innovations. The elimination of the requirement to develop everything yourself frees users, but also manufacturers, because – similar to the smartphone – almost anything is suddenly conceivable and feasible.
Every machine user can invent and build their own application with this technology. They can do this for hardware from any manufacturer, as long as they also rely on openness. And every innovative computer scientist can realize their idea of new service-based software without having to worry about the infrastructure and runtime environment themselves.
What is possible here turns the industry as we know it on its head. Collaboration based on open standards. Lively competition with previously unthinkable ideas, products and business models.
Convincing initial examples from practice
Unfortunately, by mid-2025, hardly any companies are ready to talk about pilot projects or the productive use of one of the open platforms. An attempt at an anonymous survey via the Internet has so far yielded just three responses.
None of the three companies has already gone live with a platform. But even what is known from the market shows what is now feasible across all conceivable applications and industries.
I described the first example in more detail after my visit to the Automatica robotics and automation trade fair in Munich at the end of June: Cognibotics as a partner of Bosch Rexroth and KEBA.
Here, a Swedish spin-off from Lund University’s Faculty of Technology (LTH), which has specialized in calibrating and programming the optimal movement of robotic arms for around ten years, has launched a new programming language under the name Juliet&Romeo. Built not for any proprietary operating system or specific robot manufacturer, but for any platform on which containerized apps can run.
This manufacturer of innovative robot calibration and programming did not want to have to build its own platform, but was looking for open automation platforms that were available on the market. Cognibotics first found what it was looking for at KEBA with Kemro X and at Bosch Rexroth with ctrlX OS, two of the representatives mentioned in the market overview. Others are likely to follow. In turn, these companies have found a partner in Cognibotics whose offering is far too specialized for them to want to develop the corresponding functionalities themselves. For their part, every Cognibotics customer now has the opportunity to program their own functions for the use of robots. The main thing is that they are container apps for a Linux platform.
The second example concerns Weidmüller, the last manufacturer listed alphabetically in the market overview, with its u-OS platform. The launch of the u-OS Data Hub was announced at the end of July (my message). From the new version onwards, the Data Hub is an integral part of the open operating system u-OS. It allows the reciprocal use of data from any apps running on u-OS without conversion or explicit programming. Simple configuration via parameters is all that is needed.
Weidmüller does not force any customers to use specific development and programming tools for their apps. And the use of data is not restricted to Weidmüller hardware or apps. Such openness accelerates development in the industry and is fertile ground for genuine innovation.
However, the industry must take this path courageously and say goodbye to tools and methods that have become familiar and are wrongly considered future-proof, as well as to one or two suppliers. At present, well over 90 percent of all companies are probably still using conventional OT. Proprietary, hardware-dependent, inflexible.
The openness of the platforms is also reflected in various examples of cooperation between manufacturers. There are several organizations involved in the further development of the technology. For example, the Open Industry 4.0 Alliance. In this association, a whole range of platform providers and other providers of software that also runs on Unix are working on standardizing APIs for easier data exchange, for example. Or for the better integration of security systems or other functionalities required in the automation environment.
A gift for young software developers – or a reason to quit
Not every company has experience with Linux apps. Not every company has programmers who are familiar with containerization. Above all, however, the often large, heterogeneous hardware landscapes in industrial companies are a seemingly insurmountable monster. It is difficult to decide where to start with the switch to open software. And many hardware providers are still trying to keep customers tied to their proprietary software solutions.
On the other hand, I have learned in recent years that the industry is now teeming with young software developers. They are coming out of training and are not only familiar with containerization. They don’t want to program anything else. They know that this is where the future lies.
It is becoming increasingly common for company or development management’s fear of the changeover and insistence on the old, inflexible solutions to drive such valuable employees to look for other employers. This immense danger, especially in times of a general shortage of personnel, affects automation providers as well as machine manufacturers and companies in the process industry.
The pharmaceutical industry, for example, also has major problems with proprietary control software. And waiting for a special container environment for the process industry makes no sense at all. “Not invented here” is, as usual, a bad slogan. This is because the platforms coming from the mechanical engineering sector are not limited to mechanical engineering at all. For example, Bosch Rexroth reported at a trade press conference in 2024 that up to that point, most ctrlX Core systems had not been sold for production automation, but for building automation.
The biggest advantages of open automation – like any type of composable software – can be clearly stated:
- Enormous gains in flexibility to respond to the rapidly changing and rapidly growing challenges of the markets. This applies to both manufacturers and users.
- Independence from hardware and its manufacturers allows valuable software development in the long term.
- Freedom to develop new, digital business models in the Industrial Internet of Things (IIoT).
Cybersecurity becomes feasible
As long as OT was embedded in the hardware and directly linked to it, isolation was the only security that manufacturers could offer. After all, it was not only customers and partners who had little access to this OT. The conventional proprietary control systems also offered little target for criminal attacks. Above all, however, the security solutions available in IT could not be used at all because they were built for standard software. This boundary between OT and IT is now also falling.
Suddenly, the topic of cybersecurity can also be addressed in automation. This is also urgently needed, as the Cyber Resilience Act (CRA) has been in force in Europe in particular since December 2024. With a clearly formulated timetable, it also forces almost all industrial companies to guarantee the security of manufactured goods, devices, machines and systems so that the associated OT cannot become a gateway for criminal activities.
On the one hand, open automation now makes OT a broad field for potential attacks. On the other hand, producers can access all the security tools that were previously only available to IT. And indeed, cybersecurity can become a field of application in its own right for the new automation platforms.
This is an obvious field of application, as the platform manufacturers themselves have to follow the CRA and prove that their software is secure and that all necessary measures are taken immediately in the event of an attack and corresponding software updates are installed automatically. In the market overview of smart automation platforms, there is a separate table that provides information on the current status of platform manufacturers in terms of cyber security certification.
When I started the “Composable Software” section, I had a hunch that the industry would do something important with it. The emergence of a dedicated market of open and open standards-based platforms has exceeded my expectations. The opportunities that these platforms present to revitalize the industry lovation as an innovation leader are huge. Hopefully users and manufacturers will seize the opportunity before successful competition comes from China.


