A visual pull-based inventory control method using physical or electronic cards, bins or signals to authorize production or replenishment of parts only when downstream demand consumes inventory, thereby limiting on-hand inventory and linking replenishment to actual usage.

Principle

Kanban enforces production/replenishment only in response to realized demand signals from downstream operations; the number and type of kanban tokens define maximum allowed WIP and create a feedback loop that stabilizes flow by making shortages visible and timing replenishment to consumption.

Demonstration

Illustrative scenario → A workstation uses two-card kanban for a subassembly: a production card at the supplier and a withdrawal card at the assembly cell. Recognition: when the assembly cell consumes a bin, it sends the withdrawal card to the supplier. Action: supplier produces one bin and returns the production card with the replenished bin. Consequence: inventory is limited to the number of cards, production matches consumption, and overproduction is avoided unless cards are increased.

Misapplication

Interpreting kanban as a schedule or target rather than a real-time pull signal. The error is to let cards be treated as production quotas independent of actual consumption, which converts pull into disguised push and reintroduces overproduction and inventory growth.

Consequence

Proper kanban reduces inventory, exposes process problems quickly, shortens lead time and aligns production with customer demand; misused kanban can mask upstream variability, propagate shortages if card counts are too low, or produce instability if card rules are ignored or violated.

Reversal

When demand is highly variable and replenishment lead-times are long relative to consumption variability, strict kanban may lead to frequent stockouts; in such contexts, hybrid approaches (safety stock, time-based replenishment or larger lot kanbans) or calculated decoupling inventories may be necessary.

Boundary

Clearly within: a supermarket with fixed kanban cards controlling pull between fabrication and assembly. Boundary case: electronic kanban integrated with ERP—functionally kanban but dependent on data latency and system availability. Clearly outside: forecast-driven push schedules (MPS) that release work by planned dates rather than consumption signals.

Semantic Tension

Inventory Minimization ↔ Service Level: reducing kanban card counts lowers inventory but increases risk of stockout and service-level degradation; managers must set kanban quantity to balance working capital and acceptable fill rates.

Synthesis

Kanban is a control mechanism that converts actual consumption into bounded production authority; its effectiveness depends on choosing appropriate signal granularity, card count and replenishment lead-time, and on maintaining discipline so that pull remains responsive rather than becoming a routinized schedule.