This text is a personal article by Janos Standt, Deputy Head of Public Sector, and first appeared in his LinkedIn profile (only in German). Since this is only available with a LinkedIn account, we will adopt the post here in the blog.

After the various calls for an open source offensive that have been coming up for months in connection with the digital sovereignty of Germany and the OZG (Act to Improve Online Access to Administrative Services, in short form: Online Access Act – most recently, for example, by Dr. Markus Richter in connection with innovative IT service providers (1) – I can’t help it: I have to take up the cudgels for applications that are developed model-based and can become the core of very complex heterogeneous system landscapes. After all, in the end, they also bring greater design capabilities.

First of all: Open Source Software (OSS) is good and right.

First of all: Open Source Software (OSS) is good and right. But the OSS concept alone and in general is not a solution in my view. Appropriate tools are necessary for the long-term maintenance of public authority applications. Above all, it must always be possible to implement changes due to legal requirements reliably and as quickly as possible, and this in the federal system even with local or regional differences.

Back to model-based development. But attention: commercial break 😉

In short: Model-based IT implementation – for example via a “low code” platform – provides non-IT users with tools to digitise and further develop specialist procedures & Co. themselves. At the core are easy-to-use development editors for administrative staff, with which software functions, web views and underlying processes can be built. A large part of the development work can thus be implemented quickly and efficiently directly in the specialist departments on the basis of models. In my opinion, model-based development is an opportunity for public authorities to maintain and strengthen their technical sovereignty.

What is model-based development?

In connection with model-based IT implementation, models are like the bridge between people and machines: Data models can be structured by business users and can be read directly by machines. The administrative staff in the specialist departments define what kind of logic is mapped, what data is to be captured, how it is to be validated, and what relations there are between queries and tasks. All these aspects of an application can be captured in models – and these aspects usually make up the major part of an application. Practically speaking, the models can then be used across the board: both the graphical (web) interfaces for citizens and companies (applicants) and the specialist view can use the same models.

The starting point for the model-based development of an application in the public sector is often forms. This is in the nature of our many application-oriented administrative services. Currently, of course, the OZG implementation is putting the pressure on, after all, the bells between Flensburg and Garmisch will soon be ringing in the last 24 months. However, the process can also be based on specialist procedures. Here too, model-based development makes a lot of sense in order to have the majority of the specialist application defined by the experts.

The benefits of model-based IT implementation

What I want to describe with this: Model-based IT implementation helps to ensure the sovereignty of the administration. By capturing data fields, logics and interface models for web applications and the like themselves, business managers perform a large part of the tasks in the conventional development environment. In keeping with Maria Montessori’s motif: “Help me to do it myself! This is also the case here.

Especially under the OZG time pressure, model-based development offers another decisive advantage: the development and, above all, the adaptation of applications is much faster – if only because tedious and time-consuming loops in briefing and communication between the specialist department and developers are eliminated. The technical experts design large parts of the applications in the way they make sense and are functional based on their experience.

However, model-based IT development will certainly remain relevant and helpful after 2022. Because models once developed can be reused in many different contexts. And it goes without saying that a model-based environment and the applications developed with it, like other OZG solutions, also fit into grown and complex system landscapes.

Requirements and limits of model-based IT implementation

As simple as it all sounds – admittedly: such development processes also place new demands on administrative staff. Good editors are an important building block on the way to developing applications in the specialist departments. Nevertheless, digitisation also requires a basic understanding of IT and processes, interest and a way of thinking in clear structures. A second building block is gladly a planned change management, and agile methods are of course also helpful.

Conclusion

In my view, the implementation of CCT requires a radical change in the way public authorities develop applications. The traditional route of long IT projects and exchanges between the business department and the development team simply takes too long and is not user-oriented. However, OSS alone is not the saviour, it needs meaningful concepts and the best of different worlds.

(1)  https://www.egovernment-computing.de/deutschland-braucht-innovative-it-dienstleister-a-958502/

Photocredit: Ryan Quintal / Unsplash