Optimum junction point location and inventory levels in serial hybrid push/pull production systems

The choice of a production philosophy by a manufacturing company is important. Costs for holding inventory and costs for missing due-dates on customer orders are key components of overall profitability. In this paper, an optimization model of a new, yet simple, kind of horizontally integrated push/pull hybrid production system (HIHPS) is formulated that resolves the tradeoff between these two important cost components. At optimum, it will suggest either an all-push, all-pull, or a mixed operation configuration. A mixed system would be composed of push stations followed by a cache of semi-finished product, followed by pull stations. An exact analytical solution of HIHPS is derived for the automated (low variability) case. A modification of the simulated annealing algorithm is presented to conduct the optimization search. Results of a case study are used to demonstrate the methodology and validate the model's predictions.