Lean-pull strategy for order scheduling problem in a multi-site semiconductor crystal ingot-pulling manufacturing company

Globalization has driven companies to invest in multi-site production facilities in order to improve service levels and reduce production costs. However, efficient order scheduling schemes are required to balance the load and capacity across the different sites. The present study considers the multi-site production planning problem for a real-world semiconductor wafer ingot-pulling manufacturing company owning five factories in different countries around the world. A two-phase planning model based on a lean-pull strategy is proposed. In the first phase, an available machine sends a request signal to the lean-pull logic to select appropriate orders from the central queue and accumulate them to form a production kanban. In the second phase, the fulfilled orders are assigned to the customers in such a way as to minimize the total shipping cost using a linear programming formulation. The performance of the proposed two-phase lean-pull model is compared with that of the company's existing production planning method under various workload conditions by means of numerical simulations. The results show that the proposed method achieves average improvements in the service level, average cycle time, throughput, machine utilization and profit of 6.03%, 38.48%, 22.52%, 14.26% and 12.48%, respectively. In other words, the efficiency and effectiveness of the proposed methodology are confirmed.