Development of a data space: it starts with many cups of coffee

Parties that enter into a collaborative venture to develop a data space opt for a robust and reliable infrastructure that allows only specific data to be shared with the participating parties.

At first glance, this mainly appears to be a technical and legal matter. In practice, however, it turns out that there is much more involved. At TNO Vector, we therefore took an in-depth look at the subject.

What problem does data sharing solve?

Data sharing has already produced many success stories. The advantage of a data space is that participating parties have a great deal of control over how they share their data and exactly who they give access to their data. When they do so, they can continue administrating their own data through smart privacy technology, which really is a considerable advantage.

‘Even so, this doesn’t mean that a data space is always a magic bullet,’ stresses Gijs van Houwelingen, a researcher of collaborative business models at TNO Vector.

‘Data sharing only gains added value once data applications have been developed that help solve specific problems. However, you often see at the start of a collaboration that it isn’t yet clear what problem needs to be solved. Moreover, there are often parties at the table that are in competition, and although they have common interests and are willing to collaborate, it often takes some time before they really trust each other.’

Building informal trust

Does that mean that everything should be made legally airtight right from the start? ‘If you do that, you start building formal trust, but if informal trust is still lacking at that point, it’s very difficult to reach an agreement quickly,’ notes Van Houwelingen.

‘Every detail may then lead to discussions. As a result, it may take an endless amount of time before actual data sharing occurs. Our recommendation is therefore to work on informal trust first. In other words: sitting around the table together, drinking lots of coffee, and getting to know each other better.

That, too, often requires a number of meetings, but fewer than if you tackle things formally from the outset. With a formal approach, you immediately need to spend a lot of money on legal and organisational costs, although you don't even know at that point whether a data space will actually be created.’

Starting small

Once everything is in order in terms of trust and all participating parties agree with the chosen approach, formal agreements can be concluded and signed. The key thing here is to start with a small data-sharing project and scale it up if it is successful.

An inspirational example is the HERACLES project, as part of which various parties jointly conduct research into cancer by exchanging data securely.

How do you start such a collaboration? Van Houwelingen: ‘Together, choose a specific use case that benefits all those involved and is relatively easy to implement. Work out what the minimum requirement is for this use case and keep it small and manageable at first.

A review of case studies showed that this was very clearly the common denominator of a successful approach. Often, a logical follow-up comes almost automatically after such an initial project. It helps that participants gradually gain experience in sharing data and learn in practice what the possibilities are.’

"ASML and Philips were looking for a data solution to make it easier to communicate with each other, not an expensive customised solution, so a joint data space was set up. With success." - Gijs van Houwelingen, onderzoeker, TNO Vector

High-tech companies communicating by fax

As an example of a successful case study, he cites the Smart Connected Supplier Network (SCSN). This is a collaboration in the Eindhoven region, where suppliers to companies such as ASML and Philips were looking for a data solution to make it easier to communicate with one another.

‘Sometimes they still communicated by fax, which is of course pretty remarkable in such a high-tech environment,’ Van Houwelingen jokes. ‘Then again, a customised solution would have cost a huge amount of money. Instead, they decided to set up a data space together. With success.

In this case, there was a clear problem that they could solve by sharing their data in a well-organised way. They’re currently looking at expanding their data space with a view to collaboration at the European level – a considerable scale-up.’

Things not running equally smoothly everywhere

In addition to the success stories, there are of course data space projects that do less well. However, it has proven rather difficult to uncover those stories of failure. ‘People don't like to talk about it,’ Van Houwelingen says. ‘And if they talk at all, it's more between the lines.

Even so, it’s clear that setting up a data space doesn’t go equally smoothly in every sector. For example, there are quite a few initiatives in the mobility sector with many different parties who all have different interests. This immediately makes things very complex, which often makes it extremely difficult to arrive at a data-sharing solution everyone can agree on. Here again, our advice is above all to start small and only to expand a data space further in the case of a successful start.’

Varied expertise needed

Of course, developing and managing a data space is not without costs. ‘Once talks between the participating parties start to take concrete shape, it’s therefore important to clarify quickly what investments are needed and determine together what a good revenue model could be.

All in all, there are many aspects involved. This requires having innovation specialists at the table in addition to experts in technology. Furthermore, those different experts need to remain in constant dialogue with each other to make sure that what is being built actually matches what has been agreed. This is quite difficult, because they all have a different focus and speak a “different language” because of their professional background.’

A solid foundation

To sum up, developing a data space requires very close collaboration, with the major challenge being that this collaboration has to take place between parties and specialists who would not normally interact so readily. However, if the parties manage to get the collaboration on track and carry out an actual project, this will immediately lay a solid foundation that also lends itself to future data-sharing use cases, which any new parties can join relatively easily.

Interested in getting to grips with data spaces? Feel free to contact us and find out about the possibilities.