2022 Aalto RSE report

Summary

The Research Software Engineering service allows researchers to take on more ambitious computational projects, and for existing projects to be much faster and higher quality.
About 100 projects in a bit less than two years.
Perhaps 5× return on time spent (time we spent vs time researchers saved).
There is no shortage of RSE projects, we could get more if we did more outreach (which we don’t focus on, since we would then be over-capacity).
We haven’t been able to receive much project funding, due to financial difficulties (grant rules make this difficult, we complete most projects so fast that the transaction would be too small).
We have gotten other long-term support: FCAI has supported a dedicated RSE, IT Services has provided funding to extend beyond SCI.
Our proposal is that groups receive ~1 month of free service, paid with department/school funding. After that, they should find their own funding.
- Most projects take well under one month, though, so we still focus on basic funding.
- However, there is a steady stream of longer-term project proposals which offer funding.
We ask for
- Continued basic funding at the current levels.
- Help in finding the researchers and projects who can most benefit by this service - can our results better be reported as Open Science/societal impact stories - should a small amount of RSE time be written into every grant?
- Recommendations for other schools/departments to join us.

Current status of Aalto RSE

We are part of Science-IT:

Three full-time RSEs (the only ones funded by this service)
One staff leading the RSE group + working as a RSE
Two other staff with RSE funding from specific projects
Three other staff focused on infrastructure support, but “we are all RSEs anyway”

No shortage of projects, also not yet over our capacity.
- We don’t advertise too much, since that would take us over our limit.
- Thus, there is definitely capacity to expand.

Projects fit into two main categories:

RSE projects, take days to months, recorded in our issue tracker, includes long-term support.
- The “classic work” of an RSE.
Garage help are small questions that come up in the “daily garage” and answered immediately.
- Garage is our daily support method, answering small questions.
- Garage transforms research from “trial and error” to “professional quality”.

101 researcher projects in ~ 1.75 years.
Overall “researcher time saved” is generally 5× “RSE time spent” (self-reports from customers)

From October 2020 – August 2022, about 500 visitors logged (about half of visits are recorded).
Average visit from 30-60 minutes of support.
Most are answering tech/software questions to help research, and teach people to be self-sufficient.
We also estimate overall “researcher time saved” is 5× “RSE time spent” (self-reports from customers)

Original plan: try to get most funding from grants.
Finance (for very good reasons) doesn’t want to do small transactions - minimum 1 month. Thus, we haven’t been able to accept much project funding.
- Academy/EU rules don’t allow easy internal invoicing, so must pay salary directly. This makes more overheads.
We need high-level leadership support on this topic.

Each research group gets ~1 month of free RSE time sponsored by basic funding.
After that, a group is expected to provide their own funding for future RSE projects.
However, we finish most projects in less than a month.

We should maintain at least ~2 FTE of basic funding for the near future for our current number of customers (≈ SCI).
- Any increases would be used well, though.
Future hires could be made when project funding is enough to justify costs (SCI funding as buffer between project periods)
A fair number of projects (~10-20) have written months of work into submitted grants, funded us, or offered funding.
More basic funding from other departments?
- IT Services has provided pilot funding (3 months) to expand to other schools, and has been a success.

The Finnish Center for AI has committed 4-5 years of full-time RSE funding, this was used to hire a third RSE.
We are currently (September) in planning to get more IT Services funding to secure the service beyond SCI. We will need to carefully check how this affects our staffing levels.
These type of strategic investments seem to practical and scaleable.

There is no shortage of projects, and advertising more will surely fill us up.
But, we can still increase the impact of the projects we select. Can you help point the most important projects to us?
Especially societal impact (public use of data and algorithms) could give us many more projects.

We expect this service to expand to other schools and universities in the future (bringing their own funding).
This will allow a broader knowledge base from which any individual project can draw.
Please recommend to other leaders to join us in the RSE concept.