晶圓製造廠考量製程步驟間搬運問題下之生產週期時間估算模式

Abstract

晶圓製造廠為達成高達交率、高獲利的目標，有效的掌控生產週期時間有絕對的必要性。因此，本文應用多工單等級區段基礎式週期時間演算法(BBCT-MP)，輔以本研究所發展出之車數規劃與搬運時間的估算模式，進而推估生產週期時間。 本文包含四大模組，在長程規劃模組中，估算系統所需之合適車數，使生產活動受搬運車影響而延遲的可能性降到最低；在中程規劃模組方面，承接長程規劃之車數估算結果，分析系統所需的總搬運時間，依製程步驟之順序，逐一與負荷因子或批量因子進行比對，估算各產品等級的生產週期時間；短程規劃模組中，以連批固定在製品量法制定投料法則；在現在派工模組上，本文以MIVS(Minimum Inventory Variability Schedule)派工法則為基本概念訂定合適之派車法則，以避免模擬工廠的控制不當，造成生產週期時間不必要的拉長，進而提高交期的達成度。 經實驗顯示，即使設置規劃或產品比例的改變，本文所推估的車數，可以在維持瓶頸機臺利用率的前提下，而不致造成車輛資源的浪費；在產品生產週期時間估算值與模擬結果相比較下，平均誤差皆小於3%，交期的達成度亦有98%以上之水準，故整體而言，本文所發展之生產模式，具有即時性與有效性，可作為生產規劃之搬運問題的考量上的參考依據。

To achieve the high delivery and high profit target, it’s necessary for wafer fabrication to control the product cycle time effectively. Hence, this research estimates the product cycle time with applying the Block-Based Cycle Time Estimation with Multiple-Priority Orders (BBCT-MP), and the number of overhead hoist transporter (OHT) plan and the transportation time estimation proposed in this research. This research proposed contains four modules. First, the aggregate plan (AP) module estimates the ideal number of OHT of the system to minimize the probability of production delay resulted from the OHT. Secondly, the master production schedule (MPS) module follows the number of OHT planned in the aggregate planning module, and analyzes the transportation time needed in each process step, compares the transportation time with the queuing time resulted by the loading factor or the batch factor in the order of process step to estimate the product cycle time of each priority and product. Thirdly, the detail schedule (DS) module makes the release rule with the batch-CONWIP release rule. Finally, the production activity control (PAC) module designs the vehicle management based on the concept of Minimum Inventory Variability Schedule (MIVS) in case the increase of product cycle time, and achieves a high on-time-delivery. Simulation experiment reveals that on the premise of keeping the bottleneck machine utilization, even the layout or product mix changes, the number of OHT proposed in this module won’t make any OHT be a waste. Compare the estimated product cycle time and the simulation result, the average error is less than 3%, and the achievement of on-time-delivery is higher than 98%. Consequently, the module proposed in this research is effective and efficient, and can be a feasible reference of production plan on the consideration of transportation.