台灣是全球半導體積體電路工業最大的製造基地之一,其中積體電路的封裝與測試市場占有率更是世界第一大。過去的一些投料計畫的研究,主要是針對積體電路的封裝製造工業,很少有討論到積體電路的統包(Turnkey)整合工廠的投料計畫。 在這個研究中,作者針對半導體封裝測試統包工廠,提出一個以先進產品品質計畫為基礎之投料計畫系統, 期望在有限資源限制之下,達成下列三點主要目標:(1)提升整體設備使用效能,(2)減少製造產品周期時間,以及(3)提升準時交貨率。為了平衡封裝廠與測試廠之間的產能利用率,作者使用遺傳演算法與工作負荷規律法則的啟發式研究方法,應用於投料計畫上,並藉此提升統包工廠的整體設備使用效率。 此研究提出之先進產品品質計畫為基的投料計畫系統,是用來整合知識管理系統與投料系統,以達到最佳的原物料、機器設備、與工具的選擇,並且藉由訂單的選擇最佳化,來保持大量生產的最佳效益。本研究所提出的投料計畫系統已經應用於一個半導體統包工廠,並發現比傳統策略更好的結果。
The semiconductor integrated circuit (IC) industry in Taiwan plays a significant role in the world, and accounts for a major market share of packaging and testing. Past studies have discussed job-releasing models for upstream IC manufacturing industries, assembly processes only, but only few have discussed the IC turnkey perspective. Here, the author proposes an advanced product quality planning (APQP) based job-releasing planning system (AJRPS) for an IC turnkey factory which achieves the following objectives under limited-capacity resources: (1) increasing the overall equipment efficiency (OEE) of IC test equipments, (2) reducing the cycle time, and (3) increasing the on-time delivery (OTD) rate. To balance capacity utilization between IC assembly and IC testing operations, the author uses a genetic algorithm (GA) with workload regulation (WR) heuristics for finding a better solution on job releasing and OEE maximizing in both operations. The AJRPS is integrated with a knowledge management system and job-released planning system to compose of proper solutions for material, equipment, and tools selection in a systemization way to make the selection of customers’ orders and keep mass production profitable. The proposed AJRPS was implemented in an IC turnkey factory and the improved performance was found.