Ulrich Zeh is a sales executive in the key accounts division at Mendix. Originally Dutch, the company is a global leader in low-code software platforms and was acquired by Siemens in 2018. These are certainly major points explaining why market assessments such as ABI Research (2021) on PLM software and in the Gartner 2020 Magic Quadrant designate both players as leaders. Reason enough for Detecon to inquire about potential and current challenges in the areas of application development, product design, and production planning.
Detecon: In an ideal world, there should be transparency regarding the data along the full length of the product life cycle for everyone involved. What are the current challenges facing product developers and production planners as well as the chief executives in the company? What potential does low-code offer?
Ulrich Zeh: There is a huge need across all industries to digitalize processes that, surprisingly, are still handled manually in many places, be it via media disruptions such as paper, email, or telephone or conscious decisions that are dependent on certain people. The departments want to — indeed, must — increase efficiency in this respect, but corporate IT often does not have sufficient capacity to carry out projects that will achieve these goals. Generally speaking, 80 percent of an IT department’s activities are occupied with the maintenance of existing systems. The situation leads to the emergence of classic shadow solutions that are not under the control of IT and pose security risks. Moreover, solutions that have evolved along these lines often rely on the skills of a few specific people. The use of low-code, on the other hand, can significantly increase the speed of application development while at the same time requiring fewer resources since business and IT departments can act together on joint agile teams. Compliance with company standards is assured, yet the time-to-market is shortened nevertheless.
What is Siemens/Mendix’s view here? How does Mendix complement the Siemens PLM portfolio, especially with regard to MindSphere?
Mendix is a pioneer in the low-code industry and was founded back in 2005. Even though low-code seems like so much hype at the moment, the topic is not really new. Mendix’s approach from the beginning was to democratize application development, that is, to give people who don’t have a command of any programming languages the capability to build applications. In addition to expanding capacities, this enables agile work in shorter cycles.
From Siemens’ point of view, the integration of Mendix is a sensible step enabling the enterprise itself to provide new, state-of-the-art software solutions. The portfolio components complement one another very well because Siemens MindSphere collects and consistently processes the data volume that is generated via IoT or other technologies while Mendix can make the data available as valuable information for a process, whether automatically or as appropriate to a user’s viewpoint.
What are the primary application areas for Mendix and what potential for the future do you see?
Mendix is absolutely valuable for use across all industries because very similar challenges arise wherever products are developed and produced; manufacturing is a prime example, but the same issues are found in logistics or finance, even in the public sector. In other words, wherever digitalization projects and new business models are desirable, but are hindered by low capacity and complex system access.
A classic example from the manufacturing industry is preventive maintenance, where analyzed real-time data on machine conditions might be used to map an unscheduled maintenance service that must be integrated into an overall process involving material selection, maintenance steps, and the ordering of spare parts.
Another area of application is the extension of proven core systems such as SAP, where you do not want to stray too far from the previous standard, but would still like to add new functions on top. Similarly, Mendix can be used to replace legacy or even shadow IT when much of the process knowledge has historically been contributed by a few individuals, but the platforms are at the end of their life cycle.
The motivation is similar for migrating applications to the cloud even though they are not actually cloud-capable. Since Mendix is generally based on the architecture of containers and microservices, the process logic and workflow of such applications can be newly set up on this foundation and are secure for the future. And last, but not least, multi-experience applications are also an area where Mendix can be used whenever there is an interest in making new applications accessible for voice control and other innovative communication modes.
What requirements must be met for the use of Mendix, and how long does it take to learn the programming?
A very interesting question for which there is no definite answer. The developers, as we call these people, do not have to know any programming language. Most importantly, they should know their processes inside-out and have a good idea of how the desired applications should be logically structured. Mendix has two different development environments. One is the Mendix Studio for the “Citizen Developers”, giving them the opportunity to work solely with graphics and to utilize as many reusable building blocks, pre-fabricated integrations, and starter apps as possible. The other is the Studio Pro for professional developers who want to use the full range of functions and even integrate high-code. The beauty of the structure is that both groups of people can work collaboratively on the same object so that they achieve their goals faster.
Essentially, you can produce initial, good results after just a few days of training or simply by trying things out for yourself. However, if you want to build integrations with other systems, mobile-native applications, or progressive web apps, for example, you need a little more lead-in time before you will be able to exploit the full potential of the platform.
How long do typical implementation projects such as dashboard development take?
We emphasize that a straight-forward development of an application becomes faster by a factor of 5 to 10. I could hardly believe that myself at first, but customers confirm factors of this magnitude to us. When it comes to time-to-market, however, it is important to bear in mind that the time required for the launch of a digital product entails a number of different processes, not just the development itself. Zurich Insurance created within seven days a multi-channel application that uses Microsoft Azure’s facial recognition technology to estimate people's biological age. Life insurance quotes are issued on the basis of 63,000 selfies a year. Another example is The Business Development Bank of Canada, which was able to accelerate significantly the modernization of its loan approval process; the time for application development was reduced from the expected 30 months to a mere eight months.
What are the limitations in comparison with software projects using strictly object-oriented programming languages?
One thing is clear: we do not have any doubts about high-code and its development capabilities, which will remain very important for the foreseeable future. There will still be projects during which I will consciously want to keep or realize complete individuality by using Java or .NET, for instance. And of course there are also certain limits on a low-code platform — for instance, on a completely custom display on the interface, where we use certain graphic standards. But our users always have the option of incorporating traditional high-code elements for specific purposes.
Thank you for the interview!