Normalized Systems theory: the next phase in the evolution of software development?
On October 9th TMC Belgium organised a conference on the Normalized Systems Theory, a new method of software engineering that reduces costs and improves efficiency and scalability through the automation of code generation. Professor Jan Verelst, who developed the theory at Antwerpen University, explained the basics to 75 TMC Employeneurs and clients at the Crowne Plaza hotel at airport Zaventem.
Software systems are often large and inflexible. Adaptations in the code are like stones thrown into a pond: one change leads to countless others, that all have to be manually programmed. The Normalized Systems theory addresses that problem. ‘The theory is based on software that exists of small modules,’ professor Verelst says. ‘The modules are made with automated code generators and prevent that harmful ripple effect. By regenerating the modules every three to six months the software remains young and flexible.’
NSX’s theory and technology have been applied successfully in production for years and are especially suited for information systems that use a wide variety of datasets, such as administrative systems and IoT environments that store sensor data. As a spin-off from Antwerpen University NSX was founded by Herwig Mannaert and Jan Verelst. The company develops applications based on the NS theory and has built systems for the Dutch IRS, Digipolis (a large ICT company working for the city of Antwerp) and the Helicus consortium that is developing a transport service with medical drones in Belgium.
The successful application of academic theory in the real world has pleasantly surprised Jan Verelst. ‘It doesn’t often happen that a theory developed at a university breaks through into actual software development. But the NS theory has really proven itself in practical reality. The question whether it really works hardly comes up anymore.’
Challenging
The Normalized Systems method turns out to work great for the scaling of software platforms, which has sparked interest from start-ups. ‘We’re using the code generators to build something for large scale systems and then plug in the innovative software,’ Verelst says. ‘We’ve been able to help scale a start-up that makes solar panels from 4000 to 40,000 installations in less than two years, while they employ approximately the same amount of people.’
At the conference Verelst noticed plenty of interest in his story. ‘I got a lot of detailed questions about how our technology works. That makes sense, because we’re dealing with very fundamental problems. Improvements have been announced countless times, but implementing them in practical reality is very challenging. Now we claim to have a solution that really works. So naturally people are a little skeptical at first.’
Legacy software
The aim of the conference was to introduce the NS theory to TMC’s clients and Employeneurs. ‘TMC understands the value of the NSX product because it is so innovative,’ says Employeneur Frédéric Hannes, who is doing a project at NSX and works in its R&D team. Apart from Frédéric two other TMC’ers are also working for NSX on a project basis. ‘Because TMC Employeneurs have now learnt how to work with the NS theory, TMC is able occupy a unique position in the software market,’ he explains.
Frédéric feels very much at home at NSX. ‘Everyone is very experienced and good at what they do and the working environment is relaxed. That’s due to the fact that NSX’s priorities lie less with high volume production and more with developing the theory further. They want to investigate more practical effects as well as fields of application where it can be used.’
At the conference Frédéric also noticed a lot of interest in the NS theory. ‘Most companies see how NSX’s solution is new and refreshing compared to more traditional software producers and how it can be a breakthrough in the long run. Decade old software developed with the NS theory can be modernised in a very short time.’