CMOS Image Sensor晶圓重組廠之產能規劃

Abstract

一直以來，CMOS Image Sensor晶圓重組(Wafer reconstruction, RW)廠的產能規劃都依據經驗以試算表安排產線生產；除費時之外，往往無法妥善規劃多生產區域的整體產能。對此，本研究提出一新方法，以線性規劃最佳化手法為基礎，考量相關限制，規劃每季中各產品在每週的生產數量，以及細部的機台指派。所提出的方法，除能使產能與獲利最大化外，產能規劃人員得以根據結果，進行必要的生產調整，亦可作為決策者就產能擴充與否的參考。

Capacity planning in CMOS Image Sensor wafer reconstruction (Wafer reconstruction, RW) industry includes aggregate planning and tool assignment. Aggregate planning is a quarterly decision that determines weekly production level of different products. Tool assignment is a weekly decision that matches products and tools. Conventionally, these decisions are determined based on experience. Thus the decision process is very time consuming, and the decision quality has room for improvement. The objective of this thesis is to propose a new optimization-based decision framework to resolve the problem. Considering all constraints and production parameters, new decision modules are implemented on excel solver. The new modules significantly enhance the decision quality and shorten the decision time.