標題: | 構建邏輯晶片組廠商量產初期之出貨品質評估模式 Constructing the Product Quality Evaluation System for Initial Mass Production of Chipset Industry |
作者: | 張冠達 Kuan-Ta Chang 唐麗英 Lee-Ing Tong 管理學院工業工程與管理學程 |
關鍵字: | 羅吉斯迴歸模型;量產初期;邏輯晶片組;IC設計;出貨品質;logistic regression model;mass production;Core-Logic/Chipset;IC design;shipment quality |
公開日期: | 2002 |
摘要: | 高科技產業競爭激烈,產品由研發到上市(Time to Market)所能允許的時間越來越短,公司因此缺乏充足的品質驗證時間來決定新產品的量產。倘若產品量產後才發現設計不良或錯誤時,將容易導致產品大量庫存的損失。廠商為了爭取量產初期的高獲利,尤其是在競爭對手尚未有相同產品推出前,往往願意冒著設計風險與存貨的雙重壓力,搶先量產上市。然而伴隨而來的問題,則是公司在出貨時,判斷產品良窳的檢測能力不高(即高獲利能力、低檢測能力)。因此,產品是否量產上市的決策,便考驗著高層管理者的智慧。目前業界普遍還是採取專家合議的方式為產品是否上市來背書,但這樣的方法卻隱含著人為風險,若能建立一套有科學基礎的出貨品質評估模式,將可以有效協助管理者做出產品上市與否之正確決策。本研究以新竹科學園區某邏輯晶片組IC設計廠商為研究對象,針對其在產品量產初期,以該公司歷史資料構建一個出貨品質羅吉斯迴歸模型,以提供一個快速、量化且具公信力的出貨品質評估標準,使該廠商在往後新產品要進入量產階段時,可以協助管理者做出正確量產的決策,以提升新產品的檢測能力,增加公司獲利及降低出貨風險。 Due to the highly competitive feature of the high technology industry, the lead-time of new product to the market is very short. A company usually will not have enough time to evaluate the product quality before its mass production. If there is any product designing error or defect found during the mass production, a large loss will be caused on inventory. In order to pursuit high profit in the initial production period of new product, especially before competitors announce the similar products, a company generally is willing to take the inventory risk and goes ahead for mass production. In this situation, the company will have very little time to examine the defective products. Therefore, whether or not a new product will be put in the market depends on the decision of high rank managers. This is risky since only subjective judgments of managers are involved. The main objective of this study is to construct a product quality evaluation model for the initial period of mass production of new products. The proposed method provides an efficient and reliable quality evaluation standard so that any IC design company can make the right decision on the mass production of the new products. A real case study of an IC design company is utilized to demonstrate the effectiveness of the proposed method. |
URI: | http://140.113.39.130/cdrfb3/record/nctu/#NT911031004 http://hdl.handle.net/11536/71214 |
Appears in Collections: | Thesis |