One of the most significant challenges I faced when switching back into academia wasn't living on a stipend, or volume of papers I had to read. It was trying to apply my working practices into a workplace that had no understanding or compatibility with an agile way of working. It was, for lack of a better analogy, like trying to change gears on a car without using a clutch.

Parts of the academic process can date back to the early Royal Society. While there are still areas of industry that are slow to adopt agile practices, academia is almost devoid of project management. Instead of team effort and defined scopes of work, solo heroics and gut instinct are the de facto standard. That is slowly changing, however.

I was inspired to write this post when I saw a recent nature article looking at how scrum and agile methodology are slowly bleeding into some research teams. I had two conflicting reactions to the piece.

Firstly, that this seemed like a cargo cult - with the teams simply mimicking the methodologies, rather than grokking them. When I read the testimonies, it appeared that stand-ups were used well to improve cross-team communication. However, sprints weren't really accepted, and there was no sign of any Kanban or any workload management.

But the longer I thought over it, the more I realised that perhaps the shoe simply doesn't fit that easily. Often, researchers produce knowledge, not products. Breakthroughs are sporadic, not predictable.

So, would I agree that the methodologies were implemented well in an academic setting? Probably not. But conversely, I wouldn't say that the techniques are totally unfit for purpose. Rather, there is space in the sector for some fundamental changes in their way of working. But these practices have to be a form of Agile/Kanban/Scrum/... that suits researchers and professors, not programmers and project managers.

So, my question to the community is this. What do you guys think are the most valuable practices or philosophies in an agile way of working? What sections of the process are essential to generalised project success? And what loses relevance and we step away from software development?

Personally, I suspect there's a lot of value in looking into writing requirements in an academic setting, but what are your thoughts?

Image rights:
"DSCN2607" by heyjupiter is licensed under CC BY-NC-SA 2.0