People who work with agile methods are used to visual kanban boards and how to manage them to improve workflow.
For example the kanban board below:
- To do – all items that need to be worked on
- Doing – items we are currently under work
- Done – items we have already completed the work
There is a diagram that we use to understand and act in improving the workflow. This is the cumulative flow diagram, or CFD for short.
In the CFD we are able to visualize some parameters about the workflow. We are able to visualize the correlation between the number of items being worked on and the time they take in the Doing stage.
The number of items being worked on (itens in the Doing stage) is called WIP, short for Work In Progress. Whereas the time elapsed by an item in the Doing stage is called the Lead time.
You can see both parameters — WIP and Lead time — in the CFD.
WIP on the CFD
lead time on the CFD
WIP is proportional to the average lead time. In other words, the higher the WIP, the greater the average lead time. This correlation is called Little’s Law (this correlation was described by John Little in 1961). This correlation is visible at the CFD.
In agile projects, we pay close attention to this correlation. We want to deliver fast. For this, we limit WIP to reduce the average lead time.
Now, instead of thinking about a visual kanban for work itens, let’s think about a visual kanban about the COVID-19 pandemic.
Instead of the To Do -> Doing -> Done stages, consider the following stages for the COVID-19 flow:
Never been infected -> Infected and Infecting others -> Already had and is no longer infected
Where:
- Never been infected – it is the entire population that has not yet been infected.
- Infected and infecting others – it is the population that already has the virus, whether they know about it or not. It is people who may be infecting others, until they are no longer with the virus.
- Already had and is no longer infected – it is people who have had the virus and no longer have it.
Below I share the CFD charts for two scenarios: the first with effective social isolation from week 7 and the second with effective isolation from week 3.
I am not an expert on COVID-19, but I describe below the considerations made to generate the graphs for both scenarios:
- Consider that 4 weeks after the person is infected, the person does not infect anymore, that is, the person moves from the Infected and infecting others stage to the Already had and is no longer infected stage.
- Consider a condominium of houses with a total population of 200 inhabitants.
- Consider that the expansion of the virus is exponential until isolation occurs; that is, if in week 1 there are 2 cases, in week 2 there will be 4 cases, in week 3 there will be 8 cases, and so on.
- Consider that once the isolation occurs, there are no more new cases, that is, there are no more people going from the Never been infected stage to the Infected and Infecting others stage.
Scenario 1 – effective isolation from week 7
Week – Never been infected – Infected and infecting others – Already had and is no longer infected
| week 1 | 199 | 1 | 0 |
| week 2 | 198 | 2 | 0 |
| week 3 | 196 | 4 | 0 |
| week 4 | 192 | 8 | 0 |
| week 5 | 184 | 15 | 1 |
| week 6 | 168 | 30 | 2 |
| week 7 | 136 | 60 | 4 |
| week 8 | 136 | 56 | 8 |
| week 9 | 136 | 49 | 15 |
| week 10 | 136 | 34 | 30 |
In the graph, the blue area shows, week by week, the number of people in the Never got infected stage; the brown area, the number of people in the Infected and infecting others stage. The gray area depicts the number of people in the Already had and is no longer infected stage . The blue arrow shows the moment when isolation occurs, after which there are no new cases of infection.
Scenario 2 – effective isolation from week 3
Week – Never been infected – Infected and infecting others – Already had and is no longer infected
| week 1 | 199 | 1 | 0 |
| week 2 | 198 | 2 | 0 |
| week 3 | 196 | 4 | 0 |
| week 4 | 196 | 4 | 0 |
| week 5 | 195 | 4 | 1 |
| week 6 | 198 | 0 | 2 |
| week 7 | 196 | 0 | 4 |
| week 8 | 196 | 0 | 4 |
| week 9 | 196 | 0 | 4 |
| week 10 | 196 | 0 | 4 |
If you are not familiar with agile methods, I don’t expect you to understand the theory behind this example and the CFD, but I imagine that you recognise very different results in the two graphs depending on when the WIP limit starts (blue arrow in the images above). I hope I have helped you to understand that, with social isolation, we limit WIP.
The lower the WIP (number of people currently infected and infecting others), the less time we have to get through this pandemic period.
WARNING: I am not a doctor, virologist, politician or economist. I’m just an agile specialist sharing about the CFD. This post is not intended to encourage, defend and / or go against any point about medicine, politics or economics.
[content_block id=10773 slug=cfd-en]
