In Praise of Research Software Engineers

Yesterday in the Data Innovation Research Institute we held a special event, our first ever Conference for Research Software Engineers. Sadly I was too busy yesterday to attend in person, but I did turn up at the end for the drinks reception at the end.

In case you weren’t aware, the term Research Software Engineer (RSE) is applied to the growing number of people in universities and other research organisations who combine expertise in programming with an intricate understanding of research. Although this combination of skills is extremely valuable, these people lack a formal place in the academic system. Without a name, it is difficult for people to rally around a cause, hence the creation of the term Research Software Engineer and the Research Software Engineer Association.

We have quite a few RSEs associated with the Data Innovation Research Institute in Cardiff – as you can see here. These are quite different from system administrators or other computing support staff as they are involved directly in research, working in teams alongside academics and other specialists.

One of the biggest problems facing RSEs in the UK university system is there isn’t a well-established promotions route for them. For researchers in an academic environment, performance is usually judged through publications, PhD students supervised, grants awarded and so so. Although RSEs play a vital role, especially (but not exclusively) in large collaborations, they do not usually end up as lead authors on papers and generally do not apply for grants in their own name. That means that if they are judged by these criteria they struggle to get promotion and often leave academia to work for higher pay and better terms and conditions elsewhere.

In my opinion, one of the important things that must be done to improve the lot of Research Software Engineers is to construct a career structure in parallel with the academic route  and other grades (such as laboratory technician) but judged by more appropriate criteria tailored to the reality of the job. Writing the necessary grade profiles and getting them agreed by the relevant university committees will take some time, but I think it will pay dividends in terms of better retention and job satisfaction for these highly talented people.

I hope Cardiff can take some sort of a lead in defining the role of an RSE, but this is really a national need. There are pretty uniform grade descriptions for academic and research staff across the United Kingdom so I don’t see any reason why this can’t be the case for Research Software Engineers. They are vital to many research fields already, and their importance can only grow in the future.

 

23 Responses to “In Praise of Research Software Engineers”

  1. Dark Matter Says:

    The related issue which gets less attention is the fact that many consortia uses post.doc.s to develop research software while they should be doing research. While research involves developing software but doing only software development that will be used by others is not research.

    • telescoper Says:

      Indeed. That’s an abuse of the system as it exploits early career researchers and gives them no chance of developing their career as they’re going to be judged alongside others who have more traditional `outputs’.

      • Its also arguably needed in an environment where groups cannot access money to hire people to write software.

        I suppose if there was a proper career track for this activity that might solve it….

        ….but the funding agencies seem to provide money specifically for building hardware, so why not the software to go with it? Unless they do its unclear who is going to fund any RSE jobs.

        Some universities now have in-house RSE groups – though these seem to only work on HPC algorithmic software and also appear more to be there to give advice to the people actually writing the software for experiments (rather than actually doing that). So in that respect is does not solve who would fund software people on experiments, leaving the only avenue to sneak it in via post-doc grants as is commonly done now.

    • Known Unknown Says:

      Most of the PDRAs who actually get hired for software developments are minorities and women. Many of them are graduates from non-EU countries as their degrees are not considered equivalent to the EU degrees. Some of them
      use these positions to meet nationality requirements to stay and work in the UK. Some not all, employers too use the UK nationality to poach graduates from the third world to work on these projects when they are in a premature stage. Nearer to the launch, when actual science work is done, most of these programmers are replaced with local Oxbridge graduates. Despite many initiatives to further
      careers of early career researchers in pen and paper, nothing much has actually changed and I doubt will ever change.
      There will always be people who are in high-speed career track and others who will always be in their early career stage.

      • telescoper Says:

        “Most of the PDRAs who actually get hired for software developments are minorities and women.”

        Can you quote any evidence for this assertion?

      • Known Unknown Says:

        I don’t know think any project actually publish such details.

      • telescoper Says:

        Then how do you justify the assertion?

      • Known Unknown Says:

        The same way others discard the assertion.

  2. Anton Garrett Says:

    Sadly I was too busy yesterday to attend in person, but I did turn up at the end for the drinks reception at the end.

    Glad you did your duty Peter!

  3. I was delighted to read this blog. You have accurately described much of my career. I started off my post-graduate career in 1959 by building a cosmic ray air shower experiment that used the new computer SILLIAC (a copy of ILLIAC) to analyze the observations. in 1960 I joined a small group writing Monte Carlo code to model electron-photon cascades in lead and copper absorbers. I ended up with two PhD supervisors. After two years as a postdoc in the USA learning about Radar astronomy I was inviteded back to Sydney in 1969 to join the astrophysics department under Prof B Y Mills that were building the Molonglo radio telescope. You have accurately described my career since then. In those days radio astronomers ranked behind optical astronomers in status and computer people were regarded as technicians.

  4. “One of the biggest problems facing RSEs in the UK university system is there isn’t a well-established promotions route for them.”

    But are they also on fixed-term contracts at the same stage of their careers as academic researchers are? Many people would gladly exchange a walk-on part in the war for a lead role in a cage, errm,
    a sexy but insecure job for a low-profile but permanent job. 😐 And if the job is permanent, who cares about evaluation, promotion, etc?

    • telescoper Says:

      The issue is that there is a well-defined promotion path for academic staff, with achievable criteria for career development: few people remain on a lecturer salary for their entire career. There is no such path for RSEs, which is simply unfair.

      • Just to be clear: you are talking about the career path after a permanent job has been secured, e.g. lecturer, senior lecturer, reader, professor for academics, and are advocating something similar for research software engineers?

        The question still remains, though: at what age do people in each group, on average, get permanent jobs? You mentioned system administrators. They don’t earn as much as professors (though their unplanned absence is often much more noticeable), but are often hired at an age where an academic career would mean a dozen years or so of fixed-term contracts, moving about, insecurity, etc, so some people who actually would prefer to stay in research choose to become a system administrator. Could something similar apply to research software engineers? If so, then could a career path of the type you are thinking of make this job more attractive, at least to some people, than a traditional academic path?

    • No – most of us are hired as RAs or Technical Officers, but neither group gets job security, because we’re both grant funded, so our jobs last as long as the grants do.

      • So there is no such thing as a permanently employed research software engineer?

        On the other hand, I know some such people who are paid from grant money, but have been at the same place for decades, so while not formally secure and/or permanent, in practice it can be more so than a fixed-term academic position.

      • telescoper Says:

        At the moment, most are on fixed-term positions but some people have ended up in permanent management roles after a career of those.

        Laboratory technicians and system administrators in UK universities are more likely to be on fixed-term contracts,

      • Different from some other places, where often the professor and the system administrator are the only two people in the group with permanent jobs. 😐

    • well you might care about promotion if you want to earn more money as you get older.

      • Maybe. On the other hand, most people in the world don’t, except for across-the-board rises. Many would prefer a permanent job early on, even if it means less money over the career, in return for security early on. (It might even mean more money in the sense that with a permanent job early on, one can get a loan with good conditions and buy a house, as opposed to paying rent.)

    • Anton Garrett Says:

      Hi Phillip, please would you reply on that previous thread about who played violin on your favourite recording of Bach’s unaccompanied violin music? I’m genuinely interested.

  5. Martin Says:

    So apparently I am an RSE. Good to know. It doesn’t necessarily exclude success, though; you just have to write papers as well…

    I also don’t recognise the notion that lab techs and sysadmins are on fixed-term contracts. That’s not been the case in any of the parts of the UK I’ve worked in. It is certainly the case that a programmer who doesn’t do sysadmin _or_ write papers is going to be limited in the astro jobs they can find in the UK, though the number isn’t zero.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: