Every company, organisation, public body or governmental agency today relies on data in some form – and the amount of data we all have available is increasing day by day. The same goes for the number of AI tools marketed for private as well as professional use. 

This makes the field of data science increasingly difficult to navigate – especially since many of the employees handling these data are highly skilled in their own fields but have no or little advanced data science skills. 

- I have been at Grundfos for eight years, and we see an increasing need for advanced data management skills in many different departments. Colleagues from product development over finance and to sales and marketing have access to unheard-of amounts of data – but rarely have a solid background for analysing these data and turning them into reliable decision-making support, says Mikkel Haggren Brynildsen. 

For Grundfos, one of the focal areas for increased data management is in product development, as the company’s products have turned increasingly digital – and intelligent – over the last decade. 

- We have seen how data has gone from giving us knowledge about a specific product or machine to providing us with knowledge from thousands of products or from every single part of a huge production facility – or from several similar production facilities. There is crucial knowledge residing in these data – but not that many employees are trained in understanding the potentials or complexities of them, he adds.

Which AI tools can be trusted?

And ‘complexity’ is indeed a key word here. Data today come from a wide range of sources and in many different formats – and knowing which methods and tools to use can be crucial when making investments in data management processes, the data scientist says:

- My inbox is full of proposals from companies offering the latest AI data wrangling tool. We need specialists that can make informed decisions about which tool to use in our specific context – and which are just hot air. Some companies may currently be paying hundreds of thousands of dollars for tools that are really just a piece of open-source, freely available code wrapped in a thin UI, not worth the price – a cheaper tool might give them similar results and a better ROI. This requires an entirely new knowledge and understanding of AI and machine learning tools in the context of industrial data tool procurement – and from my perspective, that means that we as companies need to offer our digital tool decision makers training in data science tools.

A key element of assessing and using such AI tools in practice is a thorough understanding of the mathematical-statistical principles behind the tools. This will enable a critical assessment to both the tools and the results they provide – a crucial step on the road to transforming data to reliable knowledge. 

A pressing need for competence development

And while the Danish universities do produce a number of highly skilled data scientists every year, these graduates rarely end up in highly specialised manufacturing companies, but rather in large companies. This is especially true for companies located far from the major cities, such as Grundfos – and exactly why a professional master such as the Master of Applied Statistics can make a huge difference, Mikkel Haggren Brynildsen says.

- It's much easier to turn a water system engineer into a water data engineer than to turn a data scientist into a water data scientist. We need access to competence development that prepares our specialists – engineers, sales representatives, product developers, design engineers and so on – for properly handling the data they have available. From having a critical approach to the tools available, to knowing how to collect, analyse and visualise data in our specific context, he says.

The key to keeping Denmark’s leading position in digitalisation

Seen in a larger perspective, strengthened data science competences will also be crucial to Denmark maintaining our position as one of the world’s leading countries within digitalisation, especially public sector digitalisation. In Mikkel Haggren Brynildsen’s view, there is a risk of disruption if companies in a wide range of sectors, as well as public authorities, do not increase their focus on utilising data.

- This is true for the individual companies – but also for Danish society at large. We are currently in a leading position where digitalisation is concerned, but we need to add a data science layer to our everyday use of digital tools and systems in order to maintain that position. We are not only in competition with companies in Germany, France, Italy, USA – but also in competition with the tech giants. If a small Danish company with a strong niche position does not add data management to their products or services, someday one of the tech giants may enter the market with a data-based competitive product or service – much cheaper or stronger than the Danish one can offer because of their size. And we all know where that story will end, he says. 

From a professional perspective, Mikkel Haggren Brynildsen also has a special focus on data management within critical infrastructure. 

- At Grundfos, we of course act within the water and water treatment sectors, which gives me a special insight into how this infrastructure works. And I see a huge potential in increased data collection and analysis – but also a need to keep the tools we use for it in Danish hands, rather than being dependent on the global tech giants. This not only goes for utilities but also for electoral systems, emergency services or communications systems, which I, personally, from a security standpoint would prefer were on Danish hands, especially in the current global situation. All this calls for increased data management competences within a wide range of sectors and industries – and that is why I think the Master of Applied Statistics comes at a perfect time in our Danish context, he finishes.