 Let's begin! We were coming to the last part of the workshop and this is the moment that I will give the microphone to Kevin Ashley, who is the chair of this Libre roundtable panel session. I will let him do the talking. I give it over to you. Mae cydwyr yw'n gweithio i'r panellys. Mae gweithio i'r cydwyr yw'r cydwyr o'r ffaith yw i'w gweld i'i gyd, i gyd yn i ddim yn gwybod i'w ddweud o'r cyfnodau. Gweithio'r cydwyr arall. Mae'r cydwyr yw'r cydwyr yw'r cydwyr yw'r cydwyr cyfnodau. Dwi'n yn gweithio'r gydwyr mewn fath. Mae'r cydwyr yw'r cydwyr e'r bodiei'n gweithio'r cydwyr yw'r cydwyr. the management of additional materials are more specifically the management and reuse of research data within UK University. So we produce training materials guidance and although our focus, because our funding is in the UK, roedd gwahbeth ffocwsidd yn yr UK iawn, mae gennym oherwydd fan i dda i fynd ymwneud mewn gwahbeth ffocwsidd. Yn ystod, bach gyddiad o'r fwrddau o'r ddau sydd ym mhangoriau yn cyflau'r wath meddwl o'r sydd yn ddiwhio ac ymwneud ymwneud isi y gymrydau'r gстиb. Mae gyddiad o'r bwysig ar y panel, gofnoddwch i'n wefiant ac i weld gan chi'n gweld, iddyn ni'n gwaith o'r cwestiynau gwahanolion oherwydd i mi yw gwahbeth yma. Gweld, yw Yw Wolfram Horstman, ychwanegwyr yma, y ddisgyrchu cyflwyno yn y Llyfrgell Llyfrgell, a'r cyflwyno cyflwyno yn y Boddlion. Rwy'n bwysig, ein cyflwyno'n ddifarton gyda'r unigol yn Blyllafoldol, ac mae'r unigol ond gyda Ymgyrch Gwyrtingon a Cyrlwn, sy'n ddifarton. Llewcas Coster, ychydig yn ddifarton gyda'r gyfeidliadol, yn y cwestiynau gwirioneddau ymlaen o'r hynny'n pethau, Cyfridsu Llanrefedd Cymru, cyfriddollawl mewn cyfriscol gyda Llanrefedd Cyfridsu Llanfannu Cymru. Mae'r residuell cyfridwyr amdellanc llei sicr enghraifft hwnnw o'r cyfridwyr sydd hefyd o'r cyfridwyr i'r cyfridwyr ein pobl yn ymwneud chi, os gallwch chi'n yn gwybod, ond arbod chi gyrfa'r cyfridwyr y bydd y bydd y bod uniond lleir wedi'u fod yn ei ddefnyddio ar gyfridwyr dyma i ddefnyddio ar gyfridwyr. Yn no rydych chi'n sicrhau William Beck, now director at TU Delft, the library is there since August 2011, background in material science and engineering, but also a career in scientific publishing, I think you worked 12 years in that area and 7 years in the library domain. So again, three quite different perspectives I think to bring to the issues of the library's role in research data management. So I'm going to be asking the panel some questions, but also going to be encouraging you to do any follow up on that, but also trying to get your opinion on some of the issues we want to deal with. And I guess I'm going to begin by asking you to reflect back. We heard a few mentions today of research life cycles or life paths of ways of decomposing the different stages that research data goes through, and we see in illustrations of a few of them. All of us will presume that the library has more of a role in some of those stages than in others. And I wonder if I could ask each of you to your own opinion of which parts of those research life cycles should the library really try hard to take a role in, and which part should the library really be stepping back from and saying in the words of one of our speakers, that is somebody else's problem? Well from, would you like to begin on that particular? Yes, that's fine. So the parts of the research life cycles libraries should engage in, well we did an interesting exercise over the last two years with the Libre Working Group on research data and we came up with 10 recommendations and I can only recommend to look at the recommendations. I think they cover most of the research life cycle or life cycles that we saw today, but I think there is some emphasis and one of the roles definitely also we saw that today or heard that today is training and advice that is maybe a scalable thing to do at a university. A role, if you would ask me to pick one role that I wouldn't recommend to take over then it's the actual active data management that is usually done by the researchers and I am skeptical that even if we reskill our library workforce significantly that we will be able actually to support that. I think that should be kept in the responsibility of the research groups themselves. But then at the end of the life cycle in the area of publication I see again a strong role for the libraries to mint identifiers or even to take over curatorial ownership for data where researchers actually leave the institution. As I am working with library systems I have a perspective from another perspective so to say. I have heard this term research life cycle but also someone mentioned called it data life cycle. I thought that is to focus too much on data because I also see that there are differences in disciplines between research at work. We have seen the example of the meteorological data. It starts with data collection but there are also other kinds of research that do not start with data collection but I think it was actually in the data life cycle slide that said conceiving. Think about what I am going to do. This is something that libraries have traditionally been doing all along because they have access to scholarly publications, the old style traditional publication articles that are again also results of research. You may have noticed that I have asked a couple of questions about linking on that level, on the old traditional level of discovery and giving access to full text. How we can on that level link from an article that is a result of research project 2 not only the data sets but maybe more the research projects information about that. Which also is a way of supporting researchers in the beginning of this life cycle. I think there has not been too much attention there. The focus has been too much on linking data sets to the publication here instead of having the broader context. If you look at the data management, in my view the storage of data is not really very interesting. You can leave it to all kinds of dedicated parties. We have seen Zinodo and there is a fixed share. The most important thing that libraries can do, I am not seeing librarians but libraries and archives maybe make the stuff findable, discoverable in a number of ways. I wonder if I can just press you a little and follow up on that before we move on to Wilma. You mentioned there that there are other actors, dance and Zinodo and Dryad and others who can potentially deal with that storage and indeed on making the stuff available. We know that some groups of researchers at least are really finding difficult to know which system to choose and I think from our researcher panel earlier on that even being able to advise people on for instance whether or not Zinodo is an appropriate choice. So I am a library director so I am not only responsible for the research data of course in the university. That is the topic of the day but I thought perhaps it was nice to bring it just one level higher. So the research life cycle is what we also use for our team that is working on research support. Research data services is a part of it. What I think are two important words is that we are not working as a library in isolation in our university. We are part of a team of supporting services so I think both for research support as a whole and research data services in specific we should be aware that we work together with people from legal services and for human resources even. So we are not isolated and that is of course true for being a library in your institute but also being part of your nation. So that is why we launch research data Netherlands to be aware that we should be aware that we are working together. So that is the other word we are working together and perhaps that is enough for now. Well perhaps there is one thing I wanted to add because you said also what we should not be doing perhaps and the idea of so it is this long tail data that is very long and very big right even perhaps more I understood then if you talk about big data. So shouldn't we for instance at the University focus on some areas if you talk about research data and then other institutes in the Netherlands like Duns but also other institutes do the same so that we together are covering the whole field. So isn't that the way forward so really focus that's. I think it's useful and I was very keen to two things you mentioned there that I think are worth as reminding yourself one about the fact that the library is not working in isolation as you say and I think we heard that point again for a number of the speakers that and okay it's unfair I phrase this question as what can the library do but we have to think of that in the context of what should the institution what's the university as a whole be doing to support its researchers and what part can the library pray in that as you say but working alongside the other professional groups the research office the IT department records managers people concerned with information management in general and also is intrigued by your phrase as well the describing this as a research data service and that accentuates the idea that is it's a service to researchers and a service to support them in what they're doing whereas management somehow can imply a role in which you're interfering and and perhaps even taking away creativity whereas in the end what we're really trying to the reason we're managing this data is to see it being reused to see research not just done once and producing one set of data but to see more research done new and different research along with the same data so are there any other views from the audience on on or do you agree or disagree with the panel about the the roles that you have yeah well I must agree with what Wilma says in our university Gant University the library's part of the research department and we think that is actually a good case because we have to work closely together with research management as a whole in the university and you see that a lot of functions that we have that there's a lot of talks between the two departments the library and research department just to get a grip on what we have to do together and who's doing what thanks enough before we go on perhaps so I can get a sense and the panel can get a sense of the room I mean I this is tagged onto a library conference and it's sponsored by library organisations so my presumption must be most of you are our library base perhaps the simplest question to ask is how many of you are not librarians okay a fair number and of those are you people who are not librarians but working in a library how many of you are not librarians and don't work in a library okay still a fair number then okay so that it's more of a mix than perhaps than than I expected so to move on then to to the panel again another we heard something about researchers needs and their expectations of you know what do they expect the university to do for them what do they expect some of the other actors in this process the publishers and others to do to help them I'd like you to think about are we well placed to meet those needs and and are the researchers expectations of us reasonable ones or are there cases where perhaps we need to manage their expectations of what others can can do for them and I wonder might begin with with you this time over and well I started to wonder the researchers panel I think also mentioned that we should be encouraging them to actually deposit the data and when we heard later talks in the training one of the also Andrew referred to it is that we should be trying to teach ourselves to be more confident to be able to approach people to be brave I always say you have three skills that you need in a library you have to be brave you have to put the user central and you have to to to actually do what you preach what you say well if you do that you will probably manage a lot and so I wondered about that because I think it is essential it's of course not the only thing they mentioned but then later on I thought in the training parts well actually that's difficult for us isn't it so how would how do we cope with that to cope specifically with with the training no to to go out and encourage and be brave and approach and not wait they asked us to do a small scale project notice not to start something big but to start it small not a way to everything is developed but just do it that's that's a different type of approach that we normally use to I think that's what they ask and that really comes but here is a room full of brave people I can see the bravery in them now you shouldn't doubt for one minute I think they're courage but go away with that that feeling so Lucas well there is a I think there is a big kind of researchers expect as far as I think research have no real expectations from libraries I have been talking to a couple nine that's true I've been talking to a couple of researchers lately because getting into this new area of research information in the broadsend not only data management and they just actually one of one of one of them at the University of Amsterdam said we I asked them do you use our our brand new discovery system no no we use Google and he said I I know that when I click on a full text link I know that it's it's goes through the library link resolver and the money that's paid by the library well the subscriptions but that's all and if we want to work together we just start a mendeley group and then we have everything there so there is it is it is it is the question what our research expectations or is the question should should we provide as libraries or provide unexpected services to like to the researchers well yes it's about both about their needs you know what what do they need us to do and what are they expecting us to do and sometimes that's the same thing and sometimes it's different if if so if they if they have other solutions out there provided by other institutions or whatever by themselves should we as libraries actually step in or not that's maybe a question to the to the audience if I may be so old I mean if to help you with you say that they have no expectations but actually I made a note of three that came out of the the session we had earlier today and okay I think this is to be fair that these are expectations they have of their university and it may the library may or may not be the best people to help with this so one is about assisting them with funder compliance more and more research funders are beginning to put rules around the creation of data management plans or deposition of data in an approved repository and these are the things that researchers are struggling with you know they know there are rules they have to comply with it's more and more rules more of the time and that's exactly what they hope their university will help them do will help them achieve that compliance in in the easiest form others talked about giving them this is not so much dealing with their own data is helping them get at other people's data giving them consistent access to data collections giving them apis as well as human readable interfaces that that's something definitely I think that that the data custodians can help with and helping them preserve the workflows that go with the data as well as the data itself so that you can truly replicate the research process that's that's that's that's that's very well but if I I see the same thing with students or something so if they don't expect anything if they're used to libraries you go there for books or something you hear that all the time although I never see a book when I where I work but if they're if they're not used to other services that we provide or could provide then they're not going to ask for it so we could probably do much more not in isolation I should say but with the larger community of the researchers themselves and the university even internationally why stick with your own local institution if researchers work internationally or based on subjects and disciplines yeah I mean to pick up one point you mentioned earlier one simple thing I thought I'm sure that that all of us could do whether it's using secant or something else is as you say researchers are simply finding stuff using Google perhaps we should make it easier for them to find not just the publications but for the data to be discoverable when when they use that and that I know that's definitely what was behind research data Australia's aim of a single data catalogue for all of Australia was simply make the data appear where people are looking for it so Wolfram finally if you can respond yeah the the the data discovery tool also ranked first in our analysis at Oxford what what has to be actually developed also for making the links between you said it was discussed too much I think it cannot be discussed too much the the links between publication and data so because data is a big big problem and unless you kind of tame the beast and one of the strategies to tame the beast is actually to say focus on those data that are underpinning publications so I think that's that is an important point but the question was a different one and so I would like to make two short comments one is things are not always what they seem so if researchers are asked what they want to have they they might give different answers so for example we did a survey in the jisc funded damaro project and they said they do not need services from the libraries but then when when asked whether they want to have data storage for example offered for free they said they would they they have things in their drawers so there would already be three petabyte in the first year was announced for being uploaded by hundreds of researchers and this was really a direct response not so this is not an extrapolated number so that's an interesting observation and the second observation is that I think researchers expectations very very very widely with with the subjects they come from and particularly to the libraries the place like oxford and the the bordlin libraries for example there is a there is a strong expectation that actually everything that is produced even digitally is held by the bordlin libraries they simply say well when when we finish the project we give it to the bordlin libraries and they will preserve it and there is it's there is very strong assumption that that actually the libraries play the role in the in the humanities and for example the medical sciences on the other hand they think well we have our stuff you know what do the libraries do anyway for us just a couple of questions I think we might ask the audience in a moment but before that do any of you have any other feelings about that issue of the researcher expectations and research needs yes just wait for the mic I don't like I don't know if it's a researcher's expectation but to make research data easier to find you have to use consistent metadata and I think researchers are always subjective about their data and so they cannot be consistent because and that's where the librarian can step in and be more objective in the description of the data so it's easier to find for other people that that's one quite specific role and I think yeah that that's that fits well with the types of skills that librarians already have or or indeed at archivists or anybody else involved in in that area so I thought perhaps before we move on to the next question do a very unscientific replication of that study of Andrew Cox's that you mentioned about it in that case we did in the UK looking at the preparedness of libraries so how many of you here feel that you've already got any sort of research data management service going in your library whether it's basic or mature or you know just something where you feel yeah we've been so very small number four five six something like that seven oh well yeah because we've heard about the one of things that happening at the rate of you but yeah so again we are and and presumably most of the rest of you you are here because you think you ought to have something or you at least want to find out how important it is for you so there's a long way to go then developing services so another simple question if you're going to be doing this we heard Wolfram said an exercise they did in a project there found that there was three petabytes of data already sitting in researchers desk drawers on flash drives on those piles of CDs that we saw Sarah mention earlier waiting to go into to university managed storage that's one of the things we encourage the DC we encourage people to do to do a simple exercise an audit to say how big is the problem you've already got how many of you know how much research data might be sitting in your institution even to an order of magnitude no i'm not surprised it's a pretty unfair question it's such a simple question to ask and it's very very difficult to do the capacity planning to know how much for instance is it going to cost us to do this if you don't even know how big the problem is if you don't know whether it's terabytes or petabytes if you don't know if there are hundreds of data sets or thousands or or millions or just five perhaps you know it makes a difference and actually finding out that it probably wasn't all that difficult for you to do was it Wolfram no no i think was a simple online um tool that was used and and it gives you again it's part of that helping it to be brave it gives you confidence when you have numbers like that you can say even if we don't like the answer at least we know what the size of the problem is i think that's that's worth doing and the response rate was high just to make it was interesting so researchers seem to be interested in the topic yeah so it's a simple thing and encourage you all to do as as as a next epic and help motivate others in the institution the decision makers to give you the resources you might need because you can begin to scope the scale of the problem even if you decide that it's not the university that in the long term is going to take care of it because we know there are lots of other places that lots of this data might end up okay a follow-up then to that about issue about the researcher needs is that there are differing views i believe on on how the extent to which one can provide a very generic set of services that suit all research domains or the extent to which you where you need support which like that provided by the subject librarians is is specific to particular research needs do you need different services for the social scientists for the material scientists for for the humanities researchers and i believe that's in in the rate at you that's an approach you take by dividing that sort of support up so i wonder if you could speak first when we're about yes of course when we started in 2008 we said okay we are lacking in the Netherlands is something like dance is that for the social for the alpha gamma sciences for the beta sciences but that's actually a two broad definition and we really heard it's now five years ago thought of where do we start then if we start such a search so we we had our three to you centers of excellence so we thought that's a good place to start because these three groups already working together so they have the actually the urge to to share data together because they're apparently working in the framework of their field in the Netherlands and but it also was too broad so we're really we're focusing now far more on climate on environmental data because if the community has settled it pretty well like an astronomy then that's fine so because the reason why we're doing this is because we want the scientific output right to be shared so we come to new innovative ideas and we need the transparency so that whole integrity issue of course so don't forget the goal there somewhere that's flying there on top that's why we're doing all these things so it's not that we want to safeguard our existence or something no we have a higher goal so if it's done well by the community fine so if it's not done well with the community and we know that and it's a community that's really close by where our subject specialists are working in then it's a good field to start with so that was sort of our approach and it's really that's also why we said dunce and three to a data center it's wise to work together because it's really specialist areas and we're not covering everything so we hope a lot of people actually join this group of research data Netherlands so that there is one entry point for the researchers for the user they can start okay so you have found that need and I wanted them we can go to wool from then next and about that the domain differences the domain differences and how much can you provide a single service that suits all research and how much do you need to to provide different targeted support for different research areas because their data needs are different yeah well I agree with the statement that was made earlier today that probably one of the great strengths of the library is that they have a subject system so for example as compared to it services they usually don't have that in an institution I'm not saying that you should compete I'm agreeing completely and we do that at Oxford as well working very very closely with IT services and research services but I think that there is a there is a strength of the libraries to actually support that subject specificity we have to do a lot of reskilling in that area to actually make the subject librarians more aware of the requirements that are that are in the individual disciplines whether or not this is sufficient for actually tackling the challenge of the the diversity of research data that that are out there is a pretty open question I think I would agree that this immersive approach of the embedded librarians is is a good way to start and to feel out how deep actually the waters are and to get a feeling how deep you can get involved in into into research data management but I would suggest to do that but in parallel to focus actually on the aspects that are not so subject specific which is discovery and and training okay so that that's worth separating out as you say there are some areas that are definitely disciplined in neutral in a sense and they're easier to look at okay so I'm not sure I'm the right person to I but I can say something I have at the University of Amsterdam we are at the moment we don't have a research data management service but we have a proposal for a program which is supposed to be a joint venture between the library and academic affairs on the university level who are about research and then the central university ICT department so this also this points to a kind of division of labor in the in that sense that's the technical infrastructure would be the domain of the ICT department whether they build something themselves themselves or maybe they will implement something that's available out there on the other hand there is already a shift in the roles of our subject libraries they are becoming more prominent and I think that's also it's very good to have them to give them a large role in their in this new area of research support um it still is within the library but in my now my my personal view which is not my library or university view so yeah is that I think it would best to have all all discipline specific stuff move away from the center to uh well to where the to where teaching and research is happening research is happening I'm intrigued I mean the point that Wolfram made about it is true that some IT services don't do that specialization but I'm aware that there are some universities at least where IT staff are often embedded within the research groups essentially to and often will come from those particular backgrounds to achieve the same ends I think as the sublover and that they can provide support that's tailored to what a particular research group needs that that builds there a connection into the very generic service that IT has and so now we have this for teaching teaching applications teaching systems that are especially decentralized IT people there but not not yet for the research if I may comment I mean this is uh maybe the truth is somewhere in the middle so I agree that this approach is I think you can find it in many places that you have IT people that are actually located in the department kind of an IT officer or so um which is I think a good approach but um they usually tend to do their own things yeah and they they build their own infrastructure so it's it's it's can't possibly happen it's it's a nice as a as a well with a little bit of CIO experience um so and the the librarians they have the central more centralized approach but they have a dialogue with the with the disciplines um an established dialogue but not in the right area so the truth is somewhere I think in in the middle between between those two so subject itarians I don't know so I guess but but for the may I but for if you look at it from the perspective of the researcher he or she should not worry about whether the services come from a library from IT or from what others and I think that's that's the main challenge for us the coming years to be able to to tackle that yeah just to give them integrated support for managing research data whether it comes helping with a grant proposal through to to managing the content I definitely agree I mean I think perhaps this is a bit given what you've all said about the early stages that most of you are at in providing research data services perhaps you don't have direct experience yet with whether it's straightforward to provide that range of supports of the disciplines but I but I wonder whether there's one I think we could ask in it and it fits in as well with how organisations in Libra could help us support this in the future one way we know that's been attractive in the UK to try and manage this is to think about consortial approaches bringing together multiple universities to work together to provide discipline support and indeed that's the approach that's been taken at dealt in in 3TU only a small number of universities are on the size and scale of Oxford and therefore can contemplate being able to provide integrated support to every single one of the disciplines that are there and for many of the rest of us the approach is taken at 3TU within the the white rose group that we heard about earlier where a group of university libraries and IT services can get together to say collectively we will support all of our researchers but perhaps one institution will concentrate on on particular areas of the sciences and other on on the social sciences and humanities perhaps you can divide it up different ways but if each of them agree to support all of the researchers they can actually provide a better quality support so how attractive is that to to any of you how yeah michael edrych from technical university of Denmark because i've been worrying a bit about this because we have a library which over the years and this is something which happened at most of the university libraries in Denmark has been centralized we don't have any department libraries anymore we don't have any subject specialists anymore so what are our connections to the the the researchers out in the departments so i'm just thinking maybe that's part of the answer to that a solution to that because we we don't have that anymore so i'm a bit worried that we you know have no chance of actually getting in in touch with the research. Sorry may i Kevin just answer go um so the division that we use in the 3TU data center is having data labs having data archive and having data services data labs is very specific tailored to research groups so there whether it's somebody from the library or somebody from our assigned center or whatever you but it's logical to have a data steward placed more or less within the group and data archive data services you have then a split between what in data services what you should do on a front office in a direct way approaching researchers and you have things that you should actually far better be sharing with a group if you talk about trained materials or or archive facilities so that's that's the steps you need to to think about maybe you don't need the library in your case for managing research support because there is you i would like to respond to that because we we've done a big information order at university and that's been been managed by the it and they only been focusing on the security of data so they only worried about are you storing data in dropbox are you using google how do you maintain make sure that it's not going into the wrong hands that it doesn't disappear in that case but they don't you know they don't care about the dissemination the sharing of data reuse and all these kind of things they have no you know they don't care about that at all so yeah i think the and that's where the library comes in okay well that's an interesting point because you say i you think you may well be right many researchers don't care about reuse but perhaps they would care more and another role that we as as institutions could have is to make them more aware of the very strong connection that's been shown in an hour of disciplines between making your data available and getting your research cited that's the sort of thing that motivates most researchers and it also motivates the universities they work in we all like to have research that's more widely cited and and there are the solid evidence that's been done in disciplines is different from economics and sell a radiology that shows this very strong link between having the data available and known to be available and linked to the publication and the fact that the research then gets more cited so perhaps that's the way we can get them interested in reuse so i wonder about and come back to that issue then just wrap it up about the the collaboration because i know collaboration can be a a difficult thing to begin in universities because in one sense we're competing with each other we compete for the star researchers we compete for students we compete for research funding and therefore finding a way to collaborate on things that are of collective interest can be quite difficult to to do it requires a bit of that bravery that that will refer to but are there any others here that would see the potential and perhaps organisation like Lieber helping you collaborate with your peers is that or am I dreaming about something here perhaps not we okay we haven't really got the enthusiasm i was struck because of course the physical library we haven't touched upon that of course at all today and you always need online and offline together so you need your visibility so visualising data but also making visual in in in an offline environment that's also a challenge we need to tackle i think see my point and it's completely different of course from the traditional libraries we used to have in our departments so it's something we are thinking also about in in Delft Netherlands and of course abroad and there was just recently i saw in the CNI announced list a new research library opened i think i'm trying to find the the name it wasn't in the states but yeah okay it's pretty big where they actually tackled that that point so how to to to create also an offline environment for the things we're doing and that's not about books anymore okay hang on i'm just thinking about a question on collaboration because my feeling is indeed that it's not always easy to get universities to volunteer to collaborate on such things most of the time something is needed as a well incentive coming from funding agencies that ask or demand a certain flow of how you work or how you disseminate things and then it can happen but i have the feeling voluntarily it's it's sometimes very difficult i mean one yeah okay michael michael day from um dcc bath um i was talking um last week to some institutions are trying to to work together and they have a lot of drivers they have a collaboration agreement they have all sorts of organizational ways of working together but even though they have much the same technical infrastructure they found it impossible to come up with some sort of idea of actually doing something concrete in this area just because the cultural differences between the institutions the way they actually use repository you know you need to be saying software so there's all these cultural impediments which make collaboration quite difficult in practice so whereas being mostly successful in terms of sharing who's like training professional training or the researchers across different institutions and those are the more soft human side of things it's been much easier to arrange rather than actually developing a shared service which would raise all sorts of questions about IT infrastructures and things okay i think that there's two um a few really interesting lessons coming out of that then about as you said michael about there are some things that are straightforward to collaborate on and some things where the where there are greater barriers and perhaps so sharing i think some more knowledge about where is collaboration straightforward and where is it difficult and therefore you'll be foolish to attempt it along with the sort of point that they made how do you make the idea of collaboration more straightforward and it seems that one way to do that might be even a little bit research to do to do a write-up of the business case perhaps behind the existing collaborations like three to you like the white rose to help us understand you know what did those institutions save in cost terms by doing this what did it cost them to do it or perhaps how much more rapidly were they able to do something as a result of that collaboration and i think some concrete lessons and measures that come out of that would might help us make the case in our own institution to say this is why the collaboration is worthwhile and i think another example from a completely different era used to be involved in in very large-scale data management i was really struck this is 25 years ago to realize that all the major oil companies who've been managing data on a very large scale for a long time that comes out of seismic surveys where they they basically blow things up and measure how the ground behaves to try and work out where the oil and gases they have one big shared data repository for all that stuff these are in companies that in other respects are intensely competitive but they realize there's no point competing over the data you can only go in and blow things up once people don't like you doing it more than once so so they share the raw data they compete in how they analyze it and they save a lot of money by having one big data store for all of that and i was very struck that if if the commercial world can identify reasons for collaboration like that in data management then we really ought to be able to do it within universities you know that's let's not persuade ourselves too much that that we can't collaborate okay so a shift then to to another strand that we heard about today around the the tools so i've been really impressed by by both the developments we heard about Zenodo but in particular the the work that Joss Whim was talking about with CCAN at at Lincoln and yet it strikes me there that I think that I suspect that's divided our audience a bit between people who look at something like that and say that's really exciting I want to try that out in in my university and other people who will look at it and say that looks frightening and difficult and technical and I hope nobody ever asked me to set anything up with that so I wonder and maybe there are other potential reactions as well how does how does it feel within your institution on taking on where you see an interesting tool like that is it easy to to adopt these things or do you think there are barriers begin as you will it's always tempting I would say um for example CCAN and and and all these nice approaches probably you were you were mentioning the boundaries we actually have in capacity to to do this and even large organization like Oxford has it so we have to decide and so we are currently on a fedora in a fedora environment and investing actually in a future fedoras program I don't know if you heard it look it up it's very very exciting because we we make it fit for research data actually um but uh the the other point that is maybe interesting with respect to tools is what what are we talking about are we talking about actual file management or are we talking about metadata management I mean we already established a little bit in the discussion the fact that um that actually not everybody can do the file management probably so but but a typical role is is metadata and I mean who who is the the traditional partner in an institution who does metadata management at industrial scale that is usually the libraries so um quite interesting is that do we really have tools for live for research data metadata management because we are mostly talking about file management I'm just leaving this as an open question well I think there's a bit of both but I'm just as interested as even if we have the tools I think it's not always straightforward for an institution to say that looks useful therefore we will provide that to our researchers because there are real costs in doing that it's yet another thing that you have to manage it's it's one thing to have an interesting research project show you wow you could do useful things with this it's quite another to turn it into a service that you can embed and commit to and not decide after a year that you've got to take it away and therefore you'll disappoint all the researchers who are going to use it I think that's a can be the hard problem and I think particularly in the area of research data management this is a very very important topic because we are making implicit assertions that we're keeping the data yeah I'm glad that you asked this question um so this uh you mentioned that my job title there is this division electronic services which will be renamed into digital services next week slightly shorter big change probably but there is also another department in this division is called DPC digital production center and they are actually already they're they're more like developers and there is currently a project internal library electronic services project going on to implement fedora as well for our institutional repository so we're replacing the old system with fedora at the same time and this is going to be interesting we will try to integrate it as much as possible with our current research information system as long as we still have that so the the idea is in this project and now we work together with my department and the other department which is also very interesting internal collaboration it's going on it's going well now this is live stream just and it's no it's no really it's it's going quite well I can say the thing is that the idea is that the metadata so this is not about research data but about the other outputs sort of the PDFs basically so the idea is that the metadata for this for this research output will be as it already is in the in the research information system it will not be duplicated again as we do now to fedora somewhere so there will be an integration using apis link linking data hopefully somehow so this will be an internal development so we and University of Amsterdam army we have the the people who could probably do the same thing but as I for research data but as I mentioned before there is this program that will be a cooperation the central university so non-library ICT well department no division I don't know it's organizational unit somehow mainly because of the the huge size probably of the of the storage needed but there might be a maybe a development possible that we will somehow but but it sounds if you do at least have the infrastructure internally to be able to to take on tools like that when it's useful to do it even if we should but no not every tool clearly but but it it's possible for you to contemplate it's not a frightening idea but I think we're also fedora so we're a fedora panel and islandora it's the line that we use and but I think that's tools more for us so I would expect if I see I'm a conference and I would a few years ago I've seen a mendele for the first time I would expect my library people to know about it and to know and understand what you can do with it you know is that a sort of question you're asking us is that what you could expect library stuff to be involved that well it sort of tools the researchers actually use for either publishing or maintaining or sharing or what have you your data so we should understand dryad and sonodo and then what have you and I think there's but there's an issue also I think for in terms of the institution supporting the researcher to yes to some extent I think you say that researchers will identify the tools that work for them in particular domains and and and it's our job not to make it difficult for them to adopt them and to make it easy for their tools to fit into the wider workflow but there are other cases where it helps for the institution to actually suggest to the researchers here are things which actually it looks as if these could be good for you and but I'm also conscious one of the jobs we have at the digital curation centre is identifying describing these things we've got a catalogue of something like 600 of these tools and and resources and I do wonder how realistic is it for anybody to know about more than a handful of those and commit to supporting them or even suggesting them it's it's and and our coverage isn't comprehensive so as long as all these tools are open and linked or linkable then you can build kind of virtual we share our knowledge base about it yeah okay very briefly because I know we're beginning to run out of time there's one more question I want to get to tools is always I always like this experience to to to quote this experience we had in in Bielefeld to stress that tools are actually not important because we did a study of one and a half years on research requirements that was for a CRIS system and repository basically and then we implemented the system together with Ghent and and Lund actually the Kathmandu system will also be presented tomorrow um within two months yeah so and that was really very a very very interesting exercise to see if you know what you need the tools are actually absolutely not not important okay that in some ways that's a positive message there okay so one last question then to wrap up so that the event today has been sponsored organized by Libre and and Openair and one of the things they're keen to know is to make sure that we actually get some lasting benefit from the workshop rather than just coming and sharing our problems and hearing about some interesting solutions they'd like to know what they can do to help those of us who are here to help our institutions move from the state we're in now which for many of us is the very early stage and and some of us even before that early stage of a basic service how can they help you establish or grow the research data services that you've got what are the right actions for Libre and Openair to take to to assist you um so although I'll begin by asking the panel this I think this is this is most definitely a question that's as much for all of you in this room as it is for the panel here who almost by definition perhaps need that help just a little bit less than the rest of us so Wilmar would you like to to begin um well of course if you look back at the day and um I think I made a remark on the training uh session I think um and and Elroy is here and referred to the foster project so that's of course coming up I think we really really meet working together and be sharing so and that should be because you asked the question here who well what about these collaboration projects and it was difficult to answer so there should be I think um um an advice or or or tools or in place to help people actually to to to to share best practices to share the training services to to to to do perhaps much more than it was ready is in place I think that might be if I look back at the day what comes as as first in my mind yes I have a very practical practical advice I think um this is from my own perspective about link linking from our discovery systems to all the other research output uh datasets and anything else we've seen slide slide presentations so um we have um we saw something presented by Wiley about from the publisher perspective so I would I think it's very important that uh international organizations like Open Air Libre and other D the Digital Curation Center and other other similar bodies exert exercise pressure on on on publishers to require researchers and authors to to provide at least links to their datasets but also preferably in my from my point of view to other information or research project information in such a way that it can be reused as a as a link open data yeah I think this is something that as as individual libraries we we can ask them but you know we just one library and if if these groups like Open Air Libre yeah that's very practical I know that very question is being asked in the UK at the moment researchers are beginning to to to ask they've got requirements and their funders that there must be a link between their papers and the data that's behind it and the data needs to be in a repository and researchers are asking is you know is there a standard for how we embed that information and quite separately as you say the problem is even if we say here is the standard for how you do it because there are some already the publishers all vary in their acceptance of it some journals will let you do it one way some will insist you do it another way some will refuse to allow it to be in there at all some will yeah so there's a lot of work to do there I think that's quite a practical thing which we can all collaborate on in and there'll be positive outcomes I just had it occur me that I discussed with Fiona earlier over lunch and there is a registry for data repositories I wondered is there also a registry for trusted data repositories because that might help well given that there's more than one way of deciding what's trusted that's indeed one of the things that prepared he is trying to deal with so there's a number of research data registries of repositories and some of them do also contain information such as whether whether repository has met some type of accreditation standard so there's the the data seal of approval the dancer that administered there was a track standard based around the work done by by RLG and others and there's an emerging ISO standard and it's in standard as well so there's even collecting that at the moment I think will be and the world data system has its own internal certification standard so many to choose from as there always is Wolfram well I think the it was already mentioned the policy development is an interesting aspect for for this organization specifically for for Libre I think but also I mean then in the European community context open air has already played I think a significant role in policy development the second point is knowledge exchange I think that well Libre as a as an organization is has this in his or her remit of course open air has established a network of national contact points through through driver already I think also this is quite a an astonishing achievement covering all Europe and building kind of a human infrastructure for knowledge exchange and getting policies and this is my third point standards actually in place maybe not developing standards themselves but enabling the institutions to implement standards just by communicating it or also by providing guidelines how to implement them so these three points okay so some work on on on standards as well there and knowledge exchange yet so there's some very concrete ones there so any other thoughts on the floor as to what Libre and open air could be doing for us to to make our lives easier to none at all it so you you can all go away and do this yourself now right I'm glad one more hi this is that's from Libre so I don't know if I am the right person to answer this question but it's just a reflection on what Libre is actually doing and how we think we are supporting libraries in Europe I completely agree with Wolfram and the knowledge exchange that Libre provides along Europe is huge actually the 10 recommendations the report about the 10 recommendations on how to start in research data is one of the most downloaded reports that we have in the web so in that really people are expecting from Libre some kind of guidance we I mean we have a steering committee is very well learned steering committees and I believe that it can be of help for libraries that are just starting in research data management or whatever that is one of the roles that I see Libre playing in Europe and absolutely policy development together with co are an opener we have been answering today so the commission is right now designing how they are how they see the research data frameworks last infrastructure in Europe and we are stakeholders their library stars stakeholders in that infrastructure and it's really good that Libre together with co are an opener and set are there informing the commission and giving raising the voice of the libraries making sure that that research data infrastructure is going to be built with the libraries yes I'm I'm wondering if it would be possible to to to help develop the curriculum for future students of the library information science schools around Europe and beyond yeah okay yeah and that's I know other organisations are working on that as well so it'd be good there's a project called dig curve that's been doing that and an international one involving people in the US around that idea of curation curriculum but I think yes absolutely worth working okay very brief very brief one through Libre and an opener we can make sure that the library voice or the voice of research institutions can be heard when we work together on this and then that's also very important to policy development okay thanks very much and I'm sure that if any other ideas occur to you that that that the the organisers of this will be pleased to hear from you offline but now unconscious time is pressing it's been a long day there's been a lot to take in and some people I know have to dash but we can just remain to thank our panel for their contributions and indeed the organisers for making the day possible thank you Kevin some final words before we rush off to the reception I have a little gift for every speaker so I'm just going to give that to them as a thank you for this day I think indeed it was a long but very very interesting day so let me give you all a little present now for the ones in the room what are they getting it's a typical gent adicus wheat for you cupardon don't ask me anything more of it it's like a little nose a pyramid form and I hope you all like it here you are so and well I hope you enjoyed reception and for the people coming to the conference tomorrow we start at 10 o'clock thank you