The research software engineering group at University of Bath is only six months old, and, so far, consists of just James Grant. But Grant has plans.
“I was taken on after the GW4, an alliance of the Universities of Bath, Bristol, Cardiff and Exeter won a bid to host a Tier 2 HPC machine. Each university agreed to contribute 0.5FTE RSE, so half of my time is spent setting up the RSE group, and the other half supporting the Tier 2 machine,” Grant says.
Grant is a strong advocate of training researchers to do their own software programming.
“The drivers for me in growing this RSE group are sustainability and reproducibility. It’s essential that any software is sustainable by the community that it’s designed for, and for that you need transfer of knowledge and skills,” he said.
Grant has two funding models in mind.
“The first is for projects with a relatively modest need, where the researchers have some software skills already. We can help develop them from ‘coders’ into software engineers, introducing tools and best practice.”
The second model is where a project lacks skills, and the level of support requires a member of the RSE group as a co-investigator.
“If a project involves some significant amount of software development, and none of the lead investigators have a track record in software, reviewers could legitimately raise concerns about their ability to deliver if they lack the ability to drive forward a significant element of the project,” Grant says.
If the RSE group can step in and help these projects to secure funding, then the university will be more likely to recognise its value and fund it, he says.
“We can engage with researchers who may not get funded otherwise, and we can help them develop good software when otherwise they might hack something together that doesn’t meet good practice. And it’s demonstrating to the university that we’re engaging with non-traditional users of computing in research.”