Science and research budgets have been feeling the squeeze since the recession. Could careful use of project management be one way to make most efficient use of these scant resources? Science and project management are not normally a happy mix – but perhaps “Agile” could help scientists manage some projects more efficiently.
Project Management And Science Are Not A Good Mix
Science thrives on creativity, flexibility, and continuous improvement of ideas and methods. These are not attributes typically associated with project management, which normally requires well defined goals and methods to be effective.
History is littered with examples of where discoveries were made by accident, in a way that could not have been planned for, such as Alexander Fleming’s discovery of penicilin. Project managers cannot plan for these discoveries, and they can completely change the direction and priorities of a research program. The presence of conventional project management would only get in the way, and hold back scientific progress.
But There Are Times When Some Project Management Could Help
Science is normally free-form and highly creative, but there are some situations in science where end goals are more well defined, and the techniques needed to achieve those goals are well known and understood.
Sometimes we have a problem which requires more formal planning
An example might be preparing a sample for an experiment. In this case we might have a good idea in advance of things like:
- deadlines for when samples are needed by
- signal strength needed when measuring the finished sample
- response of the sample to different stimuli
But conventional project management is still probably not the answer
Even though we might have a good idea these constraints and goals, we would rightly still feel nervous about committing to a strict project management timetable.
- What if something changed in the experimental setup?
- What if the sample took longer to produce than expected?
- What if we find that the signal we measure from the sample is not large enough?
If we designed, built and measured our sample in one long, project-managed process, we would probably find we had many problems at the end of it. An iterative approach makes more sense, where we build up the complexity of what we would like to achieve, and test our progress along the way.
Agile is Not Like Traditional Project Management – It Is Iterative
Agile is a project management method that came from software development, out of the need to manage complex projects, with frequently changing priorities. To me, this seems to already be a better fit to science than conventional project management.
Conventional project management takes a ‘waterfall‘ approach, where the end result is planned many months in advance, with each task and goal planned out in detail. Not very science friendly.
The emphasis of agile on the other hand is on short term bursts of activity, which are regularly reviewed against changing requirements.
This describes much better my experience of projects in science, and I am convinced that Agile project management could help in some situations.
What Might Agile Science Look Like
There are many implentations and approaches to Agile project management for science could take a number of forms. Some areas that might be important include:
- An Iterative, Evolutionary Approach. In a scientific context this might mean starting from a simple implementation of a sample or technique, and building up complexity. When preparing for an experiment, this could mean testing iterations against requirements for perhaps signal strength, or response to an input. These iterations, or time-boxes, should take place over a period of 1-2 weeks to encourage rapid feedback on what is going well, and what is not. At the end of each iteration there should be a working ‘product’ – something that would in some way achieve what the project set out to achieve.
- A very short feedback loop. A common feature in Agile methods is the ‘daily standup’ – a brief meeting that gives the whole project team the chance to update each other on what they are working towards the current iteration. This means that if one team member is struggling, or has a different idea of the plan, it is revealed immediately, rather than two or three weeks of quiet work in solitude.
- Focussing on quality. Throughout an Agile Science project the focus would be on testing the work against different quality metrics. These might be as simple as: is the sample big enough? is the signal as expected? what changes need to be made to improve things?
- Face-to-face communication. Science is often an international affair, and collaborations and consortia may span many countries or continents. Nevertheless, it is still important to favour face-to-face communication in projects. This means regularly bringing together the experiment planners, sample producers, and those who will be carrying out the experiment.
- Efficient communication. In a complex project where lots can go wrong, it is important that everyone knows the progress of different project components, and can see where there are hold-ups. A common approach is the ‘information radiator’ – essentially a wall of post-it notes showing what needs to be done, what is already done, and what is currently being worked on.