Calculating the Number of Kanban Containers


Calculating the Number of Kanban Containers

When you have determined how much buffer you require, then you are ready to calculate the final kanban container quantities . To calculate the number of containers, add together the buffer quantity and the replenishment interval, then multiply this number by the adjusted production quantity, and finally divide by the container capacity. Figure 4-18 shows the mathematical equation.

click to expand
Figure 4-18: Equation for Calculating the Container Quantity.

To determine all the container quantities, repeat this calculation for each of the production part numbers produced by the process. Figures 4-19 and 4-20 show buffer assumptions and the container calculations for the examples in Figures 4-9 and 4-10.

click to expand
Figure 4-19: Container Calculations for the Example in Figure 4-9.

click to expand
Figure 4-20: Container Calculations for the Example in Figure 4-10.



Perform a Reality Check

When your calculations are complete, the hard part is done. You now need to look at the results of these calculations and assess their reality. At this point don't let funky numbers stop the process. If you get results that look strange , then step back and look at the assumptions. Follow the steps outlined below to drive out the culprit.

When you look at the final calculations, ask yourself these questions:

  • Do the calculated quantities look like my current process?

  • Does the inventory remain the same or decrease?

If you answer No to one of these questions, then you need to review your data to determine how you arrived at these numbers. Conversely, if you answer Yes, then you should also check your data to determine whether the new quantities look is unrealistically low.

Some of the common pitfalls with initial calculations include:

  • Container sizes force added inventory

  • Incorrect production data ”production requirements, scrap, downtime, and changeovers

  • Incorrect buffer assumptions

  • Incorrect capacity assumptions

The errors associated with each of these areas revolve around using data that reflects "what we think happens" versus "what really happens."

Container Sizes

When looking at the container-sizing issue, you are trying to determine whether the containers are too big. When the containers are too big, holding parts in full container quantities gives you added inventory. In this situation, you are experiencing a rounding error. You will see this error if your calculated days-on-hand inventory is much higher than the inventory levels for the other parts . Another tip-off to this problem will be a kanban quantity that requires only one container, but increases days-on-hand inventory levels. Don't be fooled by the container number ”look at the quantity of parts being held.

To solve this problem, select smaller containers that hold fewer parts. If new containers are not an option, then look at storing less in each container to lower the quantity. If you decide to hold less inventory in the same containers, then make sure that all the containers hold the same quantity to maintain uniform counts. Also, if you select this option, consider modifying the existing carts to hold fewer parts to ensure that everyone loads the same amount.

To illustrate the impact of the containers being too big, let's revisit the container calculations in Figure 4-19. Part A's container quantities are much different from part B's. They are produced with the same mold and only differ in color . Could this be a calculation error? With a container quantity of 500, we must hold over one extra day of production. Therefore, the container for part A should be resized. Figure 4-21 shows the effects of the resized container.

click to expand
Figure 4-21: Corrected Kanban Quantities for Example 1.

Incorrect Production Data

Once you rule out containers as a culprit, determine whether you are using the correct production data. If you are using incorrect production requirements, you will either inflate or reduce the required run time. Check your current build plans and forecast. Have you used old data or artificially high requirements that drove up the production requirements? Also, review your scrap numbers for accuracy:

  • Are the scrap quantities out of date?

  • Has one scrap rate been applied in a broad brush when the different part numbers really have their own unique scrap rates?

Next , review your downtime data: Have you used real data or guessed? The downtime data will affect the time available for changeovers and can drive your replenishment interval to be negative. In the real world, a true negative replenishment interval means that you have a capacity problem. Therefore, when you calculate a negative replenishment interval, ask yourself: Am I running overtime or stocking out the customer?

The same issues apply to changeover data as for the scrap and downtime data. Are you using the real data or estimates? Did you use the same changeover times for all the parts when they are really different? Remember that since you are building a bank of time that allows you to cover all the changeovers, high changeover times result in higher replenishment intervals and higher inventories.

Incorrect Buffer Assumptions

The next item to consider is lead time. When assessing lead times ask these questions:

  • Did you use the lead time you would like to have or what you really need?

  • Do you really need all the contingencies you have planned for in the buffer inventory?

Take a good look at the buffer and make sure you are comfortable ”not extremely overweight!

Incorrect Capacity Assumptions

The last item to consider is the capacity issue. If you must produce on a six- or seven-day schedule to support your customer's fiveday schedule, then you must adjust the production requirements accordingly to account for this disconnect. Essentially, if your customer is on a daily schedule, then sum the weekly requirements and divide by six or seven to get your daily production requirements. Without scaling the requirements, you inflate the customer's requirements.

If you must produce on a six- or seven-day schedule to support your customer's five-day schedule, then make sure your buffer reflects this production disconnect. Your buffer must allow you to keep the customer supplied even though they will pull at a greater rate than you can replenish during their workweek. Finally, assess if this is a long- term situation to determine whether you need additional capacity to match your demand.

Reconciling the Problem

Take a systematic look at each of these factors to see why your calculated quantities do not parallel how you currently produce. Also, remember that the door swings both ways, so do not automatically accept your calculated kanban quantities being dramatically reduced. As a rule of thumb, review your numbers if the calculated quantities drop by more than 10 to 15 percent.

Once you have finalized the kanban quantities, then you and your team are ready to design the kanban's structure, which is the subject of Chapter 5. The rest of this chapter discusses an approach that does not utilize such in-depth calculations, looks at sizing supplier kanbans and at calculating the savings from implementing kanban scheduling.