How Many Kanbans? – Estimation Approach and Maintenance

In my previous two posts, I described how to calculate the number of kanbans (Post 1 and Post 2). However, this calculation is complex, and the result is nothing more than a very rough estimate. Hence my preferred method for determining the number of kanbans is, broadly speaking,  “just take enough, and then see if you can reduce them.” In this post, I would like to explain this approach and also discuss how and when to update the number of kanbans.

Simply Estimate the Number of Kanbans

When determining the number of kanbans, rather than going through the calculations, you could simply estimate the number of kanbans. For this you need a bit of experience, but it is doable. I also recommend not doing it alone; instead, do this as part of a group that includes a foreman or team leader from the affected shop floor.

Of course, this estimation is also not very precise. I usually go for a conservative number, where I believe I’ll definitely have enough kanbans. Having many parts instead of a few is better than missing deliveries or idling workers. But isn’t lean all about reducing material? Not all, but yes, it is one aspect of lean. Once your kanban system is up and running, of course you should verify if the number of kanbans fit the system.

Verify if the Number of Kanbans Fit the System and Adjust if Needed

How do you verify if the number of kanbans fit the system? Simple. You track your supermarket inventory. Ideally, you may have a distribution as shown below. The graph shows the inventory for one product type in your supermarket, which can vary between zero (stock out) and the maximum of all kanbans being in the supermarket. This graph below would be an ideal distribution. You never run out of stock, and since your supermarket is never full, you also never run out of work.

In reality, however, you may touch both the upper and the lower limit sometimes. Touching the upper limit is no problem. Especially if you work in larger batches, this may happen frequently. In this case your graph may look like the graph below.

Occasionally, however, you may run out of stock. As shown in the example below, your system ran out of stock once during the observed period. Of course, more kanbans could have buffered you against this stock-out, but you have to decide if this is worth the additional kanbans. If it happens very infrequently, you may decide to live with it rather than increasing your inventory.

On the other hand, if your graph looks like the graph below, then you have serious problems with your kanban system. Assuming that your system is not the bottleneck but could deliver the demand, the situation below is a sign of not enough kanbans. But by the time you get that graph, both your customer and your boss have probably already told you that they are not happy with your system.

Finally, there is also the possibility of a graph as shown below.  You always have a good inventory, and often all your kanbans are in the supermarket. In this case you may have too many kanbans. Consider removing a few kanbans out of the system to get closer to the first or second picture above. The number of kanbans you could remove altogether is the distance between the lowest inventory and zero (i.e., in the example below, about half of the kanbans are not needed). Of course, my advice would be to start slowly and remove maybe 25% of the kanbans rather than half of them. Then continue to observe and, if needed, remove more.

Through this observation of supermarket inventories, you can adjust your number of kanbans by observing your real system with real data, avoiding all those rough estimates in the kanban formula. Naturally, even if you used the kanban formula, it is advisable to verify the results through your inventory levels. If you constantly run out of stock despite your kanban calculation, then you do not have enough kanbans, regardless of what the calculation said.

Overall, a normally functioning kanban system has – very roughly – about half of the kanbans with parts in the supermarket; the other half is waiting for processing or being processed. Of course, the larger your batches, the higher your average inventory level (since you need more kanbans for the batches).

Check Occasionally or Seasonally

Production systems are constantly changing. The number of kanbans that work today may be insufficient or too much two months later. After establishing the kanban system, you need to keep an eye on it. If the customer or the system changes, you may need to update the number of kanbans. One common example is seasonality. During the season, you may need more kanbans than during the off season. In this case, add kanbans at the start of the season and remove them afterwards.

I hope this post was helpful for you. Please let me know if you have any feedback or questions. Otherwise, go out and improve your industry!