TFT LCD產業 - LCD與EPD產品組合的決定

Abstract

電子書閱讀器關鍵組成為TFT LCD產業提供之電子紙顯示器- EPD產品。在TFT LCD產業生產環境，除原有垂直多階製程、水平的多世代多廠關係外，因異質性EPD產品與FPL(Front Plane Laminate)關鍵材料，使LCD與EPD產品在各世代廠之產品組合與產能配置問題，對業者越形重要。 有鑑於此，本文針對產業環境與EPD產品的特性，設計三種規劃模組-「需求換算模組」、「FPL生產規劃模組」、「TFT 產能配置模組」，需求換算模組主要輸入庫存、預測需求、良率等生產資訊，用以換算Module、Cell後段、FPL製程階段之淨需求，作為後續FPL生產規劃模組與TFT產能配置模組之輸入值；在FPL生產規劃模組中，以線性規劃模式配置FPL產能資源，求解FPL廠各產品最適投產組合與數量，以滿足Module製程需求數量與時程；在TFT產能配置模組中，首先以線性規劃模式來建構Array自製廠區與外包廠區最適之LCD與EPD產品投產數量組合，使得外購成本、自製成本、產能閒置成本之總成本最低，完成Arrry之規劃後，再依序規劃Cell後段之產能配置，考量不同世代廠設備對不同世代玻璃支援能力，求算Cell後段製程於各廠區之投產數，使半成品與產出品庫存儲存成本以及製造成本最低，最後，亦考量Module廠區產能配置，使各廠區新購機台投資金額最小化。 實例驗證分析可知，「FPL生產規劃模組」在產能不足時可建議最佳產品投產組合，使未滿足預測需求損失金額最低化。在「TFT 產能配置模組」中，Array之LCD與EPD產品組合，受產品自製與外購相對價格「比較利益」影響，若自製EPD產品有比較利益時，將多投產EPD產品，使Cell前段產能利用率下降；Cell後段則能跨世代支援其他世代生產，使產能發揮盡致；Module投產滿足需求時，同時也兼顧產能新增與材料新購面向；上述結果顯示能達到LCD與EPD產品組合與各世代廠產能配置規劃目的。

The key component of E-book Reader Device is EPD product which is produced in TFT LCD industry. Since EPD product and FPL (Front plane Laminate) involve the original multi-generation, multi-stage, and multi-site environment, it is a complex production planning problem. Therefore, the product-mix decisions for EPD and LCD product and the capacity allocations for various generation plants become very important in the industry. Based on the complex environment and the EPD product’s particular properties, this thesis proposes three planning modules including requirement module, FPL production planning module, and TFT capacity allocation module. It is noted that the TFT capacity allocation module involves three subsystems for array, cell, and module processes. First, in the requirement module, we calculate the net demand at each stage and site in each period using related input data for other planning modules. Then, in the FPL production planning module, to fulfill the demand of module process, a linear programming model is proposed to obtain the releasing sizes and the inventory level with the objective of minimizing the inventory and shortage cost. Finally, according to information of net demand, inventory level, yield rates, related cost, margin, outsourcing vendor, cycle time, panel conversion rate between array and cell process, we construct a set of linear programming model for array, cell, and module processes in the TFT capacity allocation module. These models can obtain the optimal product-mix and the capacity allocation planning at each site in each planning period. In the array capacity allocation subsystem, the main purpose is to allocate the optimal production quantity of the EPD and LCD products for each site, and recommend outsourcing or self-production decisions as well as optimal product-mix at the same time. In the Cell capacity allocation subsystem, we provide a capacity-supporting mechanism for each generation plant with higher utilization by using capacity backup across each site. In the Module process capacity allocation subsystem, it can be provided the suggestions for the suitable purchasing quantity in order to meet the lowest machine investment cost. In addition, these capacity allocation module subsystems not only support the achievement of fulfilling demand and reducing inventory, but also meet the objective of all modules. To demonstrate the applicability of the proposed models, we present a case and perform sensitivity analysis to investigate the effect on the optimal solution and set certain parameter values (such as capacity, outsourcing price) to do what-if analysis. Furthermore, we use the comparative advantage principle to explain the change of cell utilization by adjusting the outsourcing price of EPD and LCD products. The computational results reveal that the proposed models can be provided the product-mix decisions for LCD and EPD products and capacity allocation for each stage and site in each planning period in the TFT LCD industry environment.