RSE Introductions in Wallingford

We talked briefly after the morning’s introduction session about writing weeknotes again, so I thought I’d start :)

Wallingford has been the final stop on our team’s hybrid tour of all four UKCEH sites. The Lancaster visit was the first time that all of us had the chance to meet and get to know one another in person. Months later, the two of us based in Edinburgh are lucky to be supported to visit the mothership in Wallingford. 1.5 days of back-to-back meetings, introductions to different groups and their work, it’s exhausting but fascinating! So this weeknote is more of a daynote…

Chat with the head of the Scientific Computing team. Started as a discussion of how we automate data transfer from devices in the lab whose updates are managed by vendors, in a security compliant way; ended up as a rambling tour through NetBox for asset management, ambitious but solid plans for automating the maintenance of core services as they’re transferred from Environmental Data Science into the critical but small Scientific Computing group.

A wonderful and all too brief visit to the Engineering Workshop. The folks there had a display up from a recent NERC visit so we got a curated and prepared walkthrough of recent work on custom fabrication of sensor housing for different applications (the challenges of engineering a submersible flume that will neither sink nor float to the surface!) and the history of their work stretching back nearly 50 years. We watched a watercutter course holes through a steel plate. They’re doing a lot of product fabrication in there, and that’s the thought I was left with - their skilled team is doing a lot of routine work of construction of commercial devices and it’s a valuable income stream but it comes at the cost of innovative prototyping - how do you develop an outsourcing network for fabrication of runs of 100-odd sensor+telemetry housings for different landscapes?

Both of these were side-meetings before the main event… we do a canned intro to our team to an audience that’s rich with “pockets of excellence” where there’s good practise and people in RSE roles on the teams, making assiduous effort to do reusable and reproducible work. We’re preaching to the choir here and some of the message becomes “how do we propagate the things that you do? how do we gain you the cultural support to do more work in the open, so other research groups can benefit and learn from it with minimum overhead of project communication?”

That afternoon and the next morning we had short, intense meetings with different Wallingford-based research groups, heard about their projects, dreams and challenges. A range from small groups using ad-hoc project funding to cover long-term efforts, to large groups with their own full-time software developers working on infrastructure projects that they’re building with re-use in mind, or maintaining open source modelling frameworks that have some re-use among other institutions with large RSE groups.

The range of social-technical topics was so wide, with so many places where an RSE group could engage and so relatively few of us to do it!

• A good case for a dedicated group of Research Infrastructure Engineers within Environmental Data Science specifically to work on scaling - processing, storing and publishing large volumes of data - and to work as closely as possible with the core scientific computing infrastructure team

• A frequent call for coaching and mentoring that’s half technical topics, half professional development - as an RSE group how do we create opportunities for research coders to provide this to one another?

• How do we best take advantage of our group’s position, not embedded in any specific science area, to make early-stage effort on repurposing other peoples’ work rather than do custom development ourselves?

• How do we make use of the connections in the cross-institutional RSE network that already exists, to focus and reduce effort on shared open-source projects and frameworks?

I could go on into a lot more detail of the meetings and might add to this post, but better to write something down than wait for conclusions!