Richard Darst is a Staff Scientist at Aalto University, and the organizer of the Aalto RSE group.
Abstract
Often, we hear about the goal to "start a RSE group", yet this can be a hard sell for management because of so many demands for funding and services. Instead, let's take a step back and see how we can motivate the whole picture of specilist services
Realize there are huge spaces of skills which are needed but not taught more.
Building up:
Frame in diversity: everyone needs to do work, not all have skills. Impacts different people differently read more
Dean immediately understands and gives preliminary support
Requests us to get 3 co-funded departments
Take CS, Neuroscience and Biomedical Engineering, and Applied Physics (core departments of Science-IT, no surprise there)
Remote work
C19 started before we could hire our first people
This turned out to be great: rather than being close once a week and distributed throughout the week, we were close all the time
I really wonder if we can be as efficient if things were "normal"
I have another talk on this topic I want to present sometime
Current status and way of working
Initial contact by any means (garage, personal reference, web form, email)
Gitlab for tracking issues
Divided people ad-hoc
Time-division and person-domain multiplexing, multple projects at once (not scheduling at the level of months)
We have enough projects to keep us busy while not falling too far behind. If we advertised more, we might fall behind.
Future
Funding to expand beyond school of science
More hiring
More project funding
Other notes
Funding: school (= 6-department unit) guarentees basic funding
We try to recover as much as possible from projects
Project funding
Current idea: if a project lasts more than one month, expect project funding. Otherwise basic funding.
To external funders, paying a RSE salary is
Finance doesn't like micro-transactions, which reduces ability to collect small amounts of funding
We can send invoices to anyone
Outside: huge overheads
Inside: grants don't accept it, so has to be done as salary (= much, much more admin overhead)
Cross-organization collaboration
Relation to other related service units
Science-IT (HPC group): Core part of them
IT Services: separate
Department-IT: overlapping
Research services: we take a similar high-level approach but do technical implementation as well
Other "Technical Services": rather similar in spirit
Educational Technology team:
So what is the point of RSE, and specialist services in general?
Not just knowing what to do, but doing it.
"But researchers already know how to do their work"
Open science: Researchers can do their work, but don't have time or skills to make their work reusable.
Promote diversity
Questions for discussion
What is the difference between a RSE and other university service staff?
Other services are usually aimed at all researchers at the same time. Research (software) engineers provide specialized solutions for specific research problems.
Personally I think they are somewhat similar but just supporting different aspects of research.
RSE vs expecting every researcher to know all related tools themselves
I think should be slightly different depending on the field. In fields where the computational aspects are less embedded (humanities, less-computational oriented stem) then maybe cannot be expected. In more computational oriented fields (comp physics, maths, data science) I think the line should be blurrier and RSE skills should be taught already in grad / undergrad programs
A researcher does not need to know deployment methods for software in order to do good research.
A researcher is creating prototypes, an RSE can help in bringing them closer to production level.
Researcher tools often have crude interfaces, better interfaces = more use
how do we distinguish between what users should know and what support staff should know?
Is users = researchers? Then: User knowledge, should be more field specific. Support staff knowledge should be more general (in the area where you support).
RSEs employed as a specialised group versus working in projects. Advantages/disadvantages
If you provide a service for a dedicated group, it's harder to maintain software development skills, there is less interaction with other software developers.
A lot of the skills are the same between different groups
How do we distinguish between specialist and non-specialist. should we?
There are different levels of specialization. You can be specilized in a narrow field or in a comprehensive view of a larger field.
General title: "Research Engineer"
What do you know now that you wish you had known when you started your RSE career?
How to run online events
Should RSEs also be allowed to also carry on their own research in addition to assisting others research? (assuming they want to)
How to reduce cross-organization silos?
e.g. RSE at university X needs to help university Y.
Exist outside the University! (e.g. like CSC)
Only accept funding up to a certain percentage which comes with organisational strings attached?