As global competition in the manufacturing space grows, so do corporations' needs for sophisticated and optimized management systems to enable continuous flows of information and materials across the many tiers within their supply chains. With the complexities introduced by the variability in the demand for finished goods as well as by the variability in lead-time of transportation, procurement, production and administrative activities, corporations have turned to quantitative modeling of their supply chains to address these issues. Based on the data of a heavy machinery manufacturer headquartered in the US, this research introduces a robust model for the deployment of strategic inventory buffers across a multi-echelon manufacturing system. Specifically, this study establishes a replenishment policy for inventory using a multiple bin, or Kanban, system for each part number in the assembly of products from our sponsors tractor line. We employ a numerical simulation to evaluate and optimize the various inventory deployment scenarios. Utilizing several thousand runs of the simulation, we derive a generalized treatment for each part number based on an econometric function of the parameters associated with lead-time, order frequency, inventory value and order costing. The pilot for the simulation focuses on the parts data for three earthmoving products across eight echelons, but scales to n products across m echelons. Our results show that this approach predicted the optimal quantities of Kanbans for 95% of parts to a level of accuracy +/- 3 bins.
展开▼
