By Xavier Perrin (firstname.lastname@example.org)
A lot of my missions consist in helping companies to set up a pull system. It is surprising to see how many of these companies already tried kanban, without managing to keep it going on. In these cases, talking about kanban with teams remembering the failure is always difficult: why would it work this time?
Experience shows that the reasons for the failure are almost always the same. I have identified four of them:
Lack of rigor
It is the most obvious cause of failure, but not necessarily the worst. Most of the time, the concerned managers admit themselves the lack of rigor:
- The rules have not always been properly established
- The rules have not always been properly passed on and made visible
- Line management has not been demanding enough regarding the rules
- Operators have not been properly trained (if trained at all!)
- Operators have not been involved in the setting up of the process, and have been left out of the decisions. There hasn’t been any kind of appropriation.
- Quality of production isn’t guaranteed, or adequate
No connection to the planning process
This cause is less obvious. For some, it is not obvious at all because they didn’t understand the necessity to link the kanban process to the planning process… if the planning process is efficient, that is.
First of all, it is not appropriate to go and try to set up pull process if the planning system is not efficient: kanban, no matter how strictly applied, cannot work out if capacities are not lined up with customer demand which must be anticipated by the planning process (S&OP / MPS).
Even if the planning process is efficient, there is still not to be forgotten to regularly and consistently resize the kanban loops. The loops resizing process must be defined and integrated in the planning process. Inadequacy of loop sizes and customer demand eventually lead to stock outs or excess inventories.
Kanban loops inappropriately located
Kanban loops are usually set up without any global analysis of the flows. Process steps are linked considering the apparent ease of their setting up. In many cases loops are useless. In worst situations they disturb the flows.
Pull process is not an end in itself. It is a degraded mode that is meant to be used only when a one-piece-flow cannot be created, or if processes cannot be linked with FIFO lanes.
Value Stream Mapping (VSM), executed on the shop floor with operators and other value-stream stake holders, is the only way to identify pull process opportunities: kanban loops should always come from such an analysis.
Inconsistency of goals and indicators
It is probably the most delicate aspect: the principle of pull process changes deeply the way of grasping manufacturing performance. The goal is not to keep machines running. It is to have value flowing as regularly and as rapidly as possible through the value stream, pulled from customers’ demand. If operators and supervisors are evaluated considering their process utilization there is obviously a contradiction. However, throughput time and WIP should be measured and monitored.
Most of the time, initiatives for setting up a pull system come from the shop floor or from a manufacturing manager. Thus the local control process is often not aligned with the rest of the company’s indicators. Such inconsistencies can lead managers to question the opportunity of the pull system.
Knowing these 4 classical mistakes, you can now consider using kanban successfully. Good luck!