9 Mar 2016 Amsterdam - At the e-IRG workshop in Amsterdam, Primeur Magazine had the opportunity to talk to Rene Belsø, Head of Development and International Relations of the Danish e-Infrastructure Cooporation (DeIC) about the future of e-Infrastructures. Rene Belsø explained that one needs to adapt to the changing structural setting for the infrastructure providers. A few dynamics are becoming more and more evident. For example, most average researchers, outside the realm of the more fundamental hard scientists, are pretty happy with what the commercial market can offer. They are moving to the Dropbox, to the Amazon compute Cloud and so on. They are happy with what they can get. They have a good user interface and a reliable service. Of course, they don't see the bill but sometimes they do. But otherwise it is good stuff that they can get from the private market. It has its problems with security and ownership control but still, that does not stop them. It is a development that has come here to stay.
On the other hand, the more elite researchers, from high energy physics for example, are beginning to get their science organized on a global level in a very strong way. They form communities which also attract heavy funding. For example, the ESFRI projects are funded at a level out of reach for many of the national infrastructure providers. Of course, they obviously know what they are doing. They know their infrastructure and they simply build it themselves. The consequence of this is that the national infrastructure provider is kind of losing his role and his raison d'être. He has to get refocused and getting his act together to do it in a new way. The debate going on at the moment is: Are these actually the facts? Is this how it is? Some agree and some don't. But if you do agree, what do you do about it?
What one could do, is ask why it is like this? Why are a number of researchers going to Dropbox and Amazon? This is because it works, it is better, in other words: it is a problem of quality and competence of the national infrastructure providers. What about the big ESFRI-like projects? If you want to be in bed with them, you have to be useful. You have to give them something that they appreciate. That is obviously not really happening. So again, competence is of the essence: becoming good at what you do and that fits perfectly into this e-Science debate, where one is more and more acknowledging that we need these people who are really good at understanding these e-Infrastructures as a dedicated scientific exercise without speaking about your university to help this guy. We are speaking about somebody who is a postdoc professional, a whole science field which is developed in these infrastructures. This has to evolve. What has to evolve also, is the business model, the funding structure. You cannot expect scientists to take their own funding and finance a generic infrastructure. It would be unreasonable. There has to be a different funding model also.
There has to be an answer about what to do with the Dropbox and Amazon issue. The first thing is to just stop trying to compete with them. It is ridiculous that a university tries to make its own Dropbox. It just makes a fool of itself. Usually, it is a waste of time but that doesn't mean that you should give up completely. You could work together and make add-ons. For example, the standard problem of the Dropbox is that you feel you lose controle of your data, you cannot be sure that you actually have it, and you don't know what others might use it for. End-to-end encryption could be a solution. That would be the way to co-exist.
The e-Infrastructure Reflection Group has something that is called the e-Infrastructure Commons. On the other hand, the European Commission is looking at the European Science Cloud: are these things more or less similar?Primeur Magazineasked Rene Belsø for his opinion.
First of all, all the details of the European Science Cloud are not public yet. The picture is beginning to come clear. The more Rene Belsø knows about it through the documents he has access to, the more he thinks both concepts are the same. The basic rationale is the same. You want to have low-hassle access to the tools and infrastructure types that you need. Scientists need to store and move their data. Science basically needs the same tools, some at a simple level, some at a more advanced level, but nevertheless it is the same. What you want, is a kind of basic provision of this without needing to care about how to pay, how to get access, about knowing who is in control. A researcher doesn't want to be bothered with this.
The hot issue is the funding of all this. The funding streams are supporting developments and they are distorting developments. Depending on how the money moves, you create and destroy different things. Things can be built up and they can also be blocked. There is a discussion of crystalizing to the topic that there is something wrong with the funding structure. Rene Belsø's suggestion would be to focus a little bit more on the researchers. As they get organized, you can entrust them the funding and the decision power as to which infrastructure should go the funding: how, when and why. Obviously, you cannot expect from a linguist or a theologist that they should do this by themselves but you could ask the big infrastructure projects that already exist, like the ESFRI projects, which are controlled and dominated by scientists, to form an organisation which would be in control of the strategic agenda primarily but also the funding.
If the scientists at a well-organized high level, would control the funding, that would create a very fruitful situation for developing an infrastructure. The long tail of science would not be considered in this scenario. That is not their focus. From this debate within all these ESFRI kind of projects, if earth modelling people speak with the high energy physics people, the computational chemists and so forth, a generic layer will begin to show itself. If this gets addressed and fixed, it will have a trickle-down effect on everybody else. Everybody else can just go to these Google and Apple tools.
Where there are still needs, there is a case for the infrastructure provider. There is in any case because ESFRI will just re-invent the infrastructure provider. Even outside of their scope, there is for sure a need for generic infrastructure provider.