- An approach of monitoring and regulating the flow of work in progress through a system.
- Originally came from the manufacturing process from Toyota and other Japanese companies.
- Each process is given a capacity that the team can work at and one that they cannot exceed.
- If you create a process which has the following:
- A Backlog
Sections Analysis, Development, Test are broken down into two sections of “Doing” and “Done”.
Sections Ready, Analysis, Development and Test are also given a maximum amount of tasks that they can work on at any one stage – (the maximum work in progress that they can handle).
- The only way a team can take on new work is to clear their current workload and wait for it to be moved onto the next section along the line.
- The approach works by dragging items from the previous section. E.g. Test takes its work from Development, Development from Analysis, and Analysis from Ready.
- If a section is causing a bottleneck due to work not being cleared off and therefore unable to be pulled across by the next section along, one option is to increase the work in progress limit or increase the number of resources available to that department.
- At the end of each day (or at the end of a regular time period) you then record on a graph the number of items deployed, in test, in Development and in Analysis. This allows you to monitor the changes made to the process and the effects they’ve had on the overall system
- You can record the time that a request comes into the system and how long it takes to deliver the request to the business. This allows you to work out an average time so that features can be planned into releases and timescales given to the business.