Collaborations Workshop 2018 - 2018-03-26

How to manage expectations while juggling different projects with overlapping deadlines and competing resources? - CI1-CW18

Reporter

Tania R Allard - t.allard@sheffield.ac.uk

Participants

Stephan Druskat, Robert Haines, Tania Allard, Mark Woodbridge, Stuart Grieve, Brigitta Sipocz, Alejandra Gonzalez-Beltran, Raquel Alegre

Context / Research Domain

This problem is often found within Research Software Engineering like-teams working collaboratively in a number of research projects.

Problem

Now that the necessity of having an RSE in a project is established, the issue of managing the expectations of the projects and allocating resources to multiple concurrent projects.

How to manage expectations while juggling different projects with overlapping deadlines and competing resources?

  • Difficulty in saying ‘no’
  • Difficulty of knowing when to say ‘no’ and fallout (no metrics for what happens if you say ‘no’, only if you say ‘yes’ and fail to deliver)
  • Difficulty in estimating how long a project is going to take
  • Issues of software sustainability - how is the software going to be maintained? Maintenance even on best effort basis is not easy to get into; not dealing with service maintenance
  • Differences between disciplines - collaborative, team-based disciplines (e.g. physical and life sciences) vs humanities; vision of RSEs as support rather than collaborators
  • Issue of co-location
  • The value of a project manager
    • Recognised management of people and time
  • We are competing with post-docs
  • Constraints: time, resources
  • People leaving - easier to manage in big projects where there were several people involved

Solution

Explanation of the solution to the problem you have identified

The first step is setting the expectations at the right level with the researchers.

  • Discuss with the researchers what is the MVP and how this impacts research. Then add prioritized features on top of this (until the time/money runs out). **This works really well **
    • Provide something quickly and then iterating
    • MVP produced in third of the time for the project
  • Time-focused expectations
    • Buying/selling time rather than buying/selling a finished product
    • If they want the product, they don’t have the money
  • Frequent and constant communication
  • Project management tools (even if no tool gives the perfect solution)
    • Features required: report projects, team members, time and skills
    • Harvest and Forecast (https://www.getharvest.com/forecast)- for the team
    • Tools for personal projects: Trello, Waffle, Github project management tool (Project)
  • Funders need to be made aware and accept that the running costs of a software may well keep on being incurred after the end of the project. Or “flick-of-the-switch” principle: lights go out after three years, if someone has an interest in continued use, they’ll pay you to run it.
  • Senior RSEs managing projects
  • Managing expectations all the time is not scalable
    • What we need is a culture change