“Pull” replenishment performance as a function of demand rates and setup times under optimal settings

The problem of optimizing decision variables in a single-stage replenishment loop with capacity-constrained batch processing is examined. Simulation and response surface methods are used to model total inventory and delivery performance for a continuous-review reorder point system and a single-card Kanban system. Performance tradeoff curves based on optimal settings are created using nonlinear optimization. The area under these curves is used as a single response for comparison. If tradeoff curves are experimentally replicated, main and interaction effects can also be statistically analyzed. Results show that under time-varying demand the reorder point system performs slightly better. Improvements in performance with setup time reduction are similar for both systems.