以TPM法提升半導體製造設備績效之研究

Abstract

半導體製造產業是成本與技術密集的產業，面對產品不斷推陳出新及技術世代不斷推進，半導體製造廠商必須投入大量的人力與物力提升製造設備績效來面對嚴峻的挑戰。 惟過去的半導體製造設備績效分析專注於預測模型的建立，亦或是探討設備績效相關歷史資料，分析關鍵因素，提出改善方案，但是並未指出在關鍵改善因子與問題的相關性及成本考量，並且無法指出戴明環(PDCA)改善的迴路的終點。 本研究透過資料收集與分析，將複雜的當機問題以特性要因圖(Cause and Effect Diagram)進行分類，找出組織內可改善的要因，並透過相關性分析(Correlation Analysis)、盒鬚圖(Box and Whisker Plot)及成本效益分析(Cost-Benefit Analysis)收斂出關鍵的改善因子，最後導入全面生產保養(TPM)，並以PDCA持續改善設備績效。

The past equipment performance research of the semiconductor factory focused on the historical equipment performance raw data analysis, or designing OEE forecast models. However there is less considerations among the key factor, efficiency and cost. And they didn’t address PDCA method to make close loop for performance improvement as well. In this research, we use cause and effect diagram, correlation analysis, box and whisker plot, and cost-benefit analysis to identify root cause. And using TPM method and PDCA cycle to enhance equipment performance via improving root causes.