Calculating Kanban card quantities starts with the replenishment loop: how much demand must inventory cover before replacement material arrives? The calculation converts that demand, plus an allowance for uncertainty, into container quantities. A Lean Six Sigma project also needs measurements that confirm whether the demand and replenishment assumptions match the process. (repository.lib.ncsu.edu)
Practitioners following Lean Six Sigma International’s Certificeringsprogrammer can use Kanban sizing to apply measurement and process analysis to an inventory decision. Measured operating conditions determine the initial card quantity. A trial then tests whether that quantity covers demand before the inventory limit changes.
Define the replenishment loop before counting cards
A Value Stream Mapping should show where the replenishment signal starts and where usable material returns. For the card calculation, gennemløbstid is the elapsed interval the inventory must cover, including waiting and transport inside the chosen loop. April Wagoner’s 2007 NC State University thesis distinguishes waiting from processing time in its Kanban calculations and also identifies replenishment intervals. (repository.lib.ncsu.edu)
Signal-release and material-availability timestamps give the project team a measured interval to compare with the planning-system lead time. A 2024 MIT capstone on semiconductor replenishment found discrepancies between quoted and actual delivery times and incorporated lead-time variability into its safety-stock model. For Kanban sizing, the buffer calculation needs to account for the replenishment times the team observes. (ctl.mit.edu)
The card formula separates demand coverage from protection
When one card authorises one standard container, the formula below converts demand coverage and safety stock into a card quantity. Wagoner’s 2007 review of Kanban calculations includes equivalent formulations. (repository.lib.ncsu.edu)
N = ⌈(D × L + SS) / C⌉
- N: whole cards required.
- D: demand in units per time period.
- L: replenishment time in the same time period.
- SS: safety stock in units.
- C: units authorised per container.
The ceiling brackets mean rounding upward to a whole card. Expressing safety stock as a proportion gives SS = D × L × s, so the formula becomes N = ⌈D × L × (1 + s) / C⌉. Both forms include the same allowance. Adding SS to the second formula would count the protection twice. (repository.lib.ncsu.edu)
Worked example: six cards for a 400-unit-per-shift demand
The University of Iowa’s undated Kanban Systems teaching material uses the inputs below to calculate six cards. The example is an instructional calculation; the source reports no factory outcome for it. (user.engineering.uiowa.edu)
| Kanban input | Published value |
|---|---|
| Demand | 400 units per shift |
| Order-cycle time used in the calculation | 0.2 shift |
| Container quantity | 20 units |
| Safety multiplier | 1.5 |
A multiplier of 1.5 adds a 50% allowance to demand during the stated cycle. Separating that allowance from demand coverage shows how much inventory each part of the calculation authorises:
- Demand coverage: 400 × 0.2 = 80 units.
- Safety allowance: 80 × 0.5 = 40 units.
- Cards: (80 + 40) / 20 = 6.
Six cards authorise 120 units: four containers for calculated demand and two for the safety allowance. Rounding adds no inventory in this example. The 50% allowance is a teaching assumption with no demonstrated service probability attached to it. (user.engineering.uiowa.edu)
A safety percentage does not establish a service level
NC State University’s 2011 safety-stock tutorial relates protection to uncertainty in demand and lead time. Its safety factor expresses the number of standard deviations above average lead-time demand. A percentage allowance alone supplies no measure of that variability. (scm.ncsu.edu)
For approximately normally distributed lead-time demand, the statistical calculation is SS = z × σDL, where σDL is the standard deviation of demand accumulated during replenishment. The tutorial associates z = 1.65 with approximately 95% cycle service. Cycle service measures the proportion of replenishment cycles without a stockout. The proportion of requested units supplied is a separate measure. (scm.ncsu.edu)
In a Kanban project, the service definition belongs in the analysis before the team selects a buffer. The sponsor approves the accepted shortage exposure, supported by a record of the demand history and replenishment intervals used. If the normal approximation does not fit the data, testing the proposed card quantity against observed replenishment cycles provides evidence for the buffer decision. An unsupported service percentage provides none.
Control card changes through a documented pilot
I løbet af pilottestning, stockout records need the actual signal-to-replenishment interval and the cause of each shortage. Insufficient coverage warrants a review of the card quantity. A missed signal or material awaiting quality release calls for årsagsanalyse before the team adds inventory.
The Styringsplanen assigns ownership of the card quantity and sets triggers for reassessment. It also records who monitors performance and how they respond when it departs from expectations. (leansixsigma.org) For a Kanban loop, a change in container quantity or replenishment arrangements warrants a review. Temporary extra cards need an owner and an expiry date.
Lean Six Sigma Internationals Certificeringsprogrammer progress from process and flow analysis to statistical inference and process control. The LSSx.0 pathway develops analytical skills for evaluating Kanban card quantities and the evidence behind buffer selections. (leansixsigma.org)
