Felipe Rico is a PhD candidate researching the potential of hydrogen as an alternative to fossil fuels in internal combustion engines at the Chair of Fluid Dynamics. He received the Data Champion Award for his outstanding data organization with Nextcloud and Coscine. In our interview, Felipe shares the challenges his research group faced and the RDM strategies they adopted to meet them.

RDS: First of all, congratulations again on your Data Champion Award! Let’s talk about where you started – in your group, what were data related challenges that you started out with? What motivated you to look into research data management and its tools?

FR: Well, I can speak directly from my own experience: As I started my project, I linked up with my predecessor, in part to exchange data. But because it was such a large amount of data, documentation was required for me to understand and reproduce this data – that was a huge struggle. In the end, I actually had to do a lot from scratch, e.g. to recreate and process similar data again. Hopefully, we have learned from those lessons, and now with proper documentation, my successor won't have to do it all again.

RDS: Essentially, the lack of clear documentation meant that you couldn't reuse as much data as you would have liked to?

FR: Exactly.

RDS: So how do you document your data now? How do you actually put these lessons into practice?

FR: For instance, one of our programs generates the geometry of the engine in a virtual environment. For that purpose, we use Python. Each version of this code is pushed to GitLab in order to document the development. Also, following good research practice, we make sure to create a header indicating the author and a short description, and also include instructions on how to use the program. This helps a lot for the next user to modify the inputs, to adjust them to the new requirements and reuse the program.

RDS: Your group was one of the first to systematically adopt RDM tools in your institute. What has that experience been like? Did you find it challenging? Did you see it as an opportunity to do something new and improve the old practices?

FR: In the beginning, RDM felt like a requirement imposed on us – you don’t always see the advantages immediately. As we started using RDM tools and practice more and more, some of the advantages became clear, especially for collaborative work. But some of the promised advantages have yet to pay off in practice – for instance, using Coscine to archive data. We know that in the future, Coscine will be extremely helpful to retrieve data that we created a long time ago. But for now, it just means investing time into archiving and organizing everything. Overall, it’s a discovery process and it goes step by step.

RDS: Speaking of Coscine – you received the Data Champion Award especially for your work with Coscine and Nextcloud. How do you integrate these tools into your daily research?

FR: Nextcloud has been amazing as a communication bridge between different researchers. Whenever we have a new student working on their thesis, we start by having them create an account in Nextcloud. That way, we can track their progress, retrieve the results, and properly archive their data. I also use Nextcloud on a daily basis to exchange data, papers, and results with students and with my colleagues. Coscine, as I said, is more of a long-term investment. Each time that we finish a paper or a publication for a conference, we then take the time to organize all the material and then upload it to Coscine for archiving.

RDS: Nextcloud was introduced during your research in this group. How has it changed the way you work?

FR: It has changed a lot. It happened in the middle of my research, and I can see the improvements compared to two years ago. Collaboration is much smoother now.

RDS: Could you give an example of that?

FR: When we exchanged data before, we used mostly emails – of course, the first obstacle was the capacity of the e-mail. If we exceeded 20 megabytes, we couldn't exchange data at all. We tried migrating to some alternatives like Microsoft Teams, but they lacked a clear directory structure and access control. Nextcloud made it easy to create new users, to share spaces, and to organize everything hierarchically – different folders, different files. The client itself is very useful; I have it on my laptop, on my workstation at work, and also on my workstation at home. It’s very convenient.

RDS: Your research is very collaborative. How does RDM as a group effort work for you?

FR: We have an RDM expert for the group, Sheeba Babuswamy. She’s the one in charge of sharing knowledge with us. RDM has to be learned, and every researcher is different – for instance, maybe it was easier for me to adopt RDM strategies because I very much like to have control over my data, to be tidy, to be organized. But there are different personalities and priorities in the group. I think we should treat RDM as a learning process that we need to adapt to different researchers. I don’t think there is one general solution.

RDS: Not every research group has a dedicated data manager like Sheeba. Have you felt that this has been a helpful position for your group?

FR: Of course! To have someone with the expertise and dedicated time to invest in RDM makes the world much easier. If a group can allocate resources to figuring out RDM solutions, we can achieve much better results than if we delegate these tasks to all the different researchers. We would be split between different tasks and it wouldn’t be nearly as efficient.

RDS: One of Sheeba’s initiatives was reaching out to us, the RDS team, for a data cleanup workshop for your group to discuss data organization practices for shared and individual file storage. 

FR: Yes, this workshop gave us the opportunity to start asking the right questions: How should we create a folder structure? Should it be a personal or a collective decision? How should we create metadata? What kind of metadata is important? How can we make a better metadata profile for Coscine? We realized that there are a lot of unknowns we need to find solutions for. I think it would be helpful to make this workshop a repeat event so that we do not lose sight of these questions and can find better solutions. 

RDS: Thank you for the feedback! How do you see the situation generally in your field – do you think that there is a growing awareness of the need for research data management?

FR: Yes, for sure. I'm convinced that in the future we will need better solutions for RDM and that awareness is growing. You can see it when you go to conferences – there is a lot of talk about how to exchange data, about the tools we have, how to archive data, how to create metadata and so on. It has grown exponentially. I suppose this is a direct consequence of the exponential growth of information availability: The more information we have, the more classification and management we will need – not just in scientific research, but in the industry. It’s good practice to have proper classifications of data, of inventory, of products, and so forth. That’s why I think we need to integrate RDM more strongly in education. I hope that five years from now, there will be RDM lectures and workshops at the Bachelor or Master level.

RDS: Absolutely, that’s something we’re pushing for here at the university. How about the future of your own research – is there anything that you are missing right now regarding RDM, any resources or tools?

FR: It would be wonderful to have RDM tools integrated into supercomputers. The scale of the memory capacity is completely different between a desktop, a server, and a supercomputer. That means that migrating data from the supercomputer to the desktop or to the laptop is extremely painful, especially when we are talking about data that is multiple terabytes in size. If supercomputers could integrate software for making instant backups to the servers where the data is supposed to be collected, that would make good RDM practice so much easier.

RDS: That’s a great insight, thank you! And thank you for taking the time to talk to us – congratulations again on your well-deserved award.

FR: Perfect. Thank you so much.