Written by Carlos Martinez-Ortiz and Florian Huber

Facts

What: First Conference for Research Software Engineers (RSE) in Germany
When: 4-6 June 2019
Where: Albert Einstein Science Park in Potsdam

But wait… what is a RSE anyway?

Like many other areas of society, the field of scientific research currently sees a quickly growing influence of new digital technologies and new IT concepts. Not surprisingly, this results in an ever-increasing need for people to tackle the increasing digitalization of research.

In the past, people in many different positions within the academic landscape would take care of developing digital tools such as research software. Typically this included many PhD students and postdocs, hence positions with deviating job profiles and non-permanent contracts.

In view of the rising need for digital expertise and the missing adequate positions and job descriptions within most academic systems, an initiative to establish a dedicated profession of the Research Software Engineer (RSE) was started.

deRSE19 aerial group photo (CC-BY Antonia Cozacu, Jan Philipp Dietrich, de-RSE e.V.)deRSE19 aerial group photo (CC-BY Antonia Cozacu, Jan Philipp Dietrich, de-RSE e.V.)

This initiative started in the UK with the UK Research Software Engineer Association (see also the not-so-brief history of RSEs). Several countries followed and founded RSE organizations including the Netherlands (NL-RSE) and Germany (deRSE).

As Alys Brett pointed out during her keynote, RSEs sit somewhere in between a researcher and a software engineer. The term RSE is still somewhat vague as it covers a broad spectrum of expertise and activities; there are no universal job specification or career paths defined for RSEs, although some initiatives are emerging. Many people may be a RSE without realizing it!

Slide from the presenation of Alys Brett pointing at the position of RSEs in between Software Engineering and Research.
Slide from the presenation of Alys Brett pointing at the position of RSEs in between Software Engineering and Research.

If we think of RSEs as individuals which combine (in different proportions) software engineering skills & scientific research skills then it indeed covers a very broad range. This means the RSE community is in itself very diverse, which has the advantage of diversity of knowledge and expertise. However, the downside of this diversity is that it is difficult to put a RSE in a box and clearly identify who is a RSE.

RSE demographics

So, is this just a small group of people? Not really — since 2016, an international survey is conducted, asking individuals questions regarding their education, field of expertise, technical skills and demographic information. During the conference, Stephan Janosch and Martin Hammitzsch used this survey to give an impression of how the RSE landscape looks like in Germany. A similar analysis was conducted at DLR, although some people at DLR do not identify as RSEs but as researchers which do software (see talk abstract).

In general, the RSE landscape looks very similar to the current situation in the Netherlands. It still is a long way to go to establish the role of RSEs within the academic landscape. One indicator is the large fraction of non-permanent contracts among RSEs (see bar plots below). Another one is that actual job titles in Germany still show the lack of dedicated positions related to research software engineering (see word clouds below).

It at first seems as if RSEs in the Netherlands can already rely a better job situation (more permanent positions) and better fitting job titles. However, bare in mind that the number of respondents in Germany was higher than in the Netherlands: 333 vs 54 which may induce some bias in the statistics. The numbers also heavily depend on who considers herself/himself a RSE. Hence who is actually responding to the survey. In the Netherlands we believe that a large number of participants working at the eScience Center dominate the statistics.

Plots from (or based on) a [2018 RSE survey](https://github.com/softwaresaved/international-survey/tree/master/analysis/2018). A larger fraction of RSEs in the Netherlands holds a permanent position (bar plots on the left. Permanent RSE employment: 57% vs 25% in Germany). And RSE-related job titles are still not common in the German academic setting (Job title wordclouds on the right. Top right: Netherlands. Bottom right: Germany). Both aspects will hopefully change through the actions of the RSE initiative. Also be aware that statistics for the Netherlands were fairly low for the survey and are likely to be dominated by people from the Netherlands eScience Center.Plots from (or based on) a 2018 RSE survey. A larger fraction of RSEs in the Netherlands holds a permanent position (bar plots on the left. Permanent RSE employment: 57% vs 25% in Germany). And RSE-related job titles are still not common in the German academic setting (Job title wordclouds on the right. Top right: Netherlands. Bottom right: Germany). Both aspects will hopefully change through the actions of the RSE initiative. Also be aware that statistics for the Netherlands were fairly low for the survey and are likely to be dominated by people from the Netherlands eScience Center.

What will the future bring for RSE’s

We don’t know — undoubtedly research software will continue to play an important part in science and with it the role of RSEs will become more prominent.

What we do know for sure is that deRSE was a fantastic conference and a great success! And it reminded us that would be very beneficial for the RSE community in the Netherlands to have another local RSE conference (we had a first meet-up last year) where RSEs from different institutions can come together and share their experiences and thoughts. Join the NL-RSE mailing list for updates on upcoming RSE events in the Netherlands.

Some impressions from the conference

Over the three days of the conference we saw a lot of interesting presentations, posters, attended workshops and discussion sessions. Clearly too much to do all contributions justice. Here we just picked a few highlights that stuck with us and were not mentioned above:

  • The secret behind the growth of RSE Groups in the UK (see their blog post).

  • Andreas Zeller gave a great entertaining keynote lecture on Sustainable Research Software. His suggestion was to use literate programming as well as prototyping. *Literate programming+ Prototyping = Literate prototyping = Notebooks *(see also the recent post on literate programming by Johan Hidding)

  • There were about 40 posters and most of them were very briefly (1 minute) pitched in a special lightening session. This made it very clear how broad the audience was in terms of backgrounds and scientific research fields.

Presenters of brief poster pitches during the lightening session.Presenters of brief poster pitches during the lightening session.

Final notes

Many of the presentations were recorded and will soon be made available. Best to check out updates on their twitter channel or the conference website. Attending this conference was a lot of fun, so thanks for organizing and see you next time!

Any comments, suggestions, or questions?
Get in touch!
Carlos on twitter
Florian on twitter