Following the digital e-IRG Conference under Croatian EU Presidency, that took place online May 25-26, 2020, we had the opportunity to invite Erik Fledderus from SURF and e-IRG delegate for the Netherlands, in our virtual studio. We talked about how partnerships can help to organise national e-Infrastructures, looking at the organisation of horizontal and vertical e-Infrastructures and Cloud organisations on a national level.
We saw a presentation from you at the e-IRG Workshop where you explained about partnerships and how these could be used or seen as a model for e-infrastructure organisations. Can you tell a little bit more about this?
Erik Fledderus: Yes, the theme of the workshop was partnerships and it very much started off as a discussion on the partnerships that are going to happen around the European Open
Science Cloud. My focus was slightly more from the point of a Member State, in the sense of what type of partnerships are occurring there, and one step back, what is a partnership actually solving. What is it trying to solve? Taking that as a starting point, I discussed a number of building blocks that at the moment occur within the Netherlands and that can help us to build two types of partnerships: the horizontal type between national & regional or local e-infrastructure facilities, resources and services, and the vertical type between e-infrastructure facilities, resources and services & scientific user communities.
Partnerships is a general term. What is typically a partnership? What are the important elements of a partnership?
Erik Fledderus: A partnership in my opinion is something that tries to solve a number of questions on a level that is beyond the daily user and the daily interactions. Without this partnership the users or the service providers at some point run into situations that are difficult to solve. Quite often they are very creative and try to solve this in an 'ad hoc' fashion but as always, this has its shortcomings. I think a partnership can try to solve it in a more structural way. A key element of a partnership is that it tries to solve a very concrete issue from daily practice in a more structural way.
One thing is what I call a vertical partnership. Many of the infrastructures and the infrastructure services that are provided by an organisation like SURF and similar ones in other countries, are there for users that may be scientists or scientific communities. They may be national, they may be quite often international, and thus cross-border. As you go to the daily practice of what scientists and researchers try to solve, then you see that, besides the size, especially in the use of these infrastructures, they are quite advanced. So, it takes quite some time to get it really into their field. One of the issues that they really have to deal with is how is it funded: "Do I have to cater for that within my project budget or is someone else taking care of that?"
Are you now talking about the organisation of the national e-infrastructures and the national funding?
Erik Fledderus: Well, it happens both in a national but also in an international situation. Suppose this situation: you have a large cooporation with your colleagues in a specific field that may be a European project but it may also be organized without already being a European project. You work with colleagues from all across Europe or maybe from all over the world. You want to have a collective facility to store data, to compute on the data, to visualize results, to collaborate on papers and to share your results and your data. The open science community or the open science movement really aims for those types of interactions.
The next question is: "Do we have to cater for that ourselves or is there something out there, outside our community, outside our interaction that we can use?" Of course, quite
often commercial off-the-shelf facilities are well-known and we use them also in our daily lives but they are often not really useful for the advanced things that the scientific community wants to aim at. If that is the case or the question is: "It is there but it is very expensive", how is this being resolved? This is not only happening within a Member State but also across Member States.
On the European level, one is talking about the European Open Science Cloud, EuroHPC which are some very big initiatives. How should the connection be between local or national initiatives and the European Open Science Cloud? Will that be dictated by a few big countries with big national organisations or by the smaller ones or is there some flexibility? How do you see that developing?
Erik Fledderus: I don't know whether it will be dictated by countries. These are big questions but these are also questions that are there already for a long time. How would you resolve that within a country? Instead of starting directly from the European Open Science Cloud (EOSC), start from a National Open Science Cloud (NOSC). How would you resolve this issue? This vertical type of partnership that you would like to establish but also its counterpart, a horizontal type of partnership where you talk about the infrastructures on a national scale but also on a more local scale within a university or within an institute.
You see that especially when it comes to data. We don't want to have a very centralized approach where all the data is centralized, stored somewhere in a country or on a continent, but rather use a hybrid, federated type of approach. The various instances need to interact smoothly. Starting from a National Open Science Cloud point of view, what type of regulations, what type of partnership would you like to establish? What is possible? Every Member State has its flavours. I think the National Nodes report from e-IRG also shows that. I really have to go back to what is possible within the Netherlands and see what type of partnerships and rules around these partnerships I can export to the discussion table of EOSC. These are, I would say, the things that a Member State can put on the table of the Executive Board when it comes to how to arrange such a partnership on a European scale.
Coming back to your question: "Will that be an issue of the big states, will the big Member States dictate how that will be possible?" I would say, look at how many different regulations are needed within the various Member States, and then come up with a few that have shown to be quite effective and try to interact between these ways of operations. This means that we much more need to focus on what effective ways of operations we have in the various Member States instead of big countries dictating their rules.
Yes, dictating indeed is a very strong word of course but I meant dominating which is also a strong word. There clearly is a lot of discussion needed before we decide and make things getting to work. How do you see the role of the e-IRG in that?
Erik Fledderus: I think the e-IRG should definitely continue its work when it comes to getting grip on the various ways a Member State organizes itself without prescribing how it should work. A first step is making transparent how things are organized, why things are organized the way they are. If you want to transgress from there, you could advise Member States how their colleagues and their neighbours are organizing themselves.
In the past five to ten years, you see definitely Member States learning from each other. I have been quite active in Scandinavia, in Sweden and Finland, and also within Germany a few times. There you see that the discussion on how you want to organize this is really taking into account expertise from neighbouring countries. There is really quite an active looking over the fence and learning from each other. I think e-IRG can really stimulate that by taking this information and broadening it. It is not about just a few spots that are clear and present and where one is active in communicating their strategy but also taking in from other Member States that are not that active in communicating, their information. That is how you really get it forward. That is a really important step and a really important role that is also observed by the Executive Boards from the EOSC.
A next step could be - and I am careful with being very strong on that - taking all the ways things are now arranged within the various Member States and see what would be a top four or top five of ways to arrange it, depending on your starting position. We know indeed that changing things involves a change of many actors and that will be difficult. This will take time. You shouldn't be too overoptimistic by saying: "We have just one way of arranging it and you all have to comply. That is not realistic. So, first we have to see what is the top four or top five of ways of arranging it.
We are looking forward to these discussions and what comes out of it. Thank you very much for this interview.