??????? ???? Design and Implementation of an

Document Sample
??????? ???? Design and Implementation of an Powered By Docstoc
					        知識管理與創新 實務報告
Design and Implementation of an Intelligent
   Manufacturing Execution System for
  Semiconductor Manufacturing Industry
           IEEE ISIE 2006, July 9-12, 2006
           Volume: 4, pages: 2948-2953


  指導老師:楊東麟 教授
  第二組:張國華、王建閎、蔡沅澄、李綺玟、田 杰、彭正儀
      林維均、陳建甫、林子健、吳明潁、黃家遠、劉毅宏

                                             報告:張昱仁
Agenda
•   What Is MES
•   What Is an Intelligent MES
•   The Benefit to a FAB
•   Case Study
•   Summary
•   Q&A
What Is MES
• AMR(Advanced Manufacturing Research):
   – An integrated architecture for plant wide information management that
     groups applications and functions around a central common database
     used to share product and process data among the applications.
• MESA(MES association):
   – MES delivers information that enables the optimization of production
     activities from order launch to finished goods.
• Process Management
   – flow / service / tools / recipe / EDC plan....
• Equipment Management
   – State Mgnt. / AT / UT / PM
• Quality Management
   – EDA / SPC / Sampling
• WIP(Work In Process) Management
   – Track In / Out / Hold / Release / split / rework ….
Objective of MES
•   Reduce cost
•   Cut cycle time
•   Improve quality
•   Increase equipment utilization
•   Guarantee on time delivery
    What Is an Intelligent MES
•   Discover the newfound knowledge
•   Massive data in system could be more meaningful
•   Support high-level decisions making
•   Speed up the creation of new strategy
•   Control and manage equipment
•   Automate fab processing
•   Flexibly alter material flow
•   Enable real-time analysis
How To Design an Intelligent MES
• Integration of MES and data warehouses
• Integration of data warehouses and
  decision-making analysis
• Integration of decision-making analysis
  and data mining systems
• Set up of a data mining engine
Integration of MES and data
warehouses
•   Collection
•   Transformation
•   Loading an indexing
•   Quality control check
•   Announcement or publication
•   Renovation
•   Search
•   Checks and preparations
Integration of data warehouses and
decision-making analysis
•   Designing a schema for data warehouse
•   Designing a fact table
•   Designing the dimension table
•   Designing a multidimensional data model
    – Time
    – Area
    – Product
    – Quality
Star Schema
Integration of decision-making
analysis and data mining systems
Procedures of the intelligent MES
Training data from the quality database
The decision tree for training data
The Benefit to a FAB


                           Improved
                           Unimproved


                       Yield 5.87% ↑
                       CT 23.49% ↓
Case Study - Productivity Improvement
• FAB產能增加, 目前的機台數與使用情形, 無法滿足未來的
  產能需求 , 需再購置新機台. 但 Cost = ?

                  買新機台
       現況 26K
                                   目標 28K


  K
                              K
  26                          28

  24                          26

                              24
  22
                              22
  20
        現況                    20
                                    未來
Case Study - Productivity Improvement
• Issue
   – 因買新機台成本消耗太大, 所以在經由團隊討
     論後發現,若能提升機台的UT(utilization) 與
     POH(Pieces Of an Hour) 將能突破這瓶頸.
• Action
   – 成立QIT (Quality Improvement Team)
QIT Procedure

 •   主題選定
 •   現況分析(需求)
 •   要因分析
 •   真因(需求)驗證
 •   真因(需求)再驗證
 •   對策擬定與實施
 •   效果確認
 •   標準化與水平展開
 •   標準化後效果確認
 •   效益分析
QIT 主題選定
     因買新機台成本消耗太大, 所以若能提升 chamber UT 與
 POH 將能突破這瓶頸.

                                        QIT
                     Improve Chamber UT and POH

                                              28K
         CH
          A                K

    CH
     B
              CH
               C
                      28         26K
                      26

                      24

                      22

                      20
                               Before         After
Case Study - Productivity Improvement
   由現況分析針對 chamber UT loss做出魚骨圖
                   機台別


     Serial Mode
                           idle time limit
    OHT 派工不及
                            Chamber Type
                            different
                                             Chamber UT
                                                Loss
  無法control派
  貨順序
                         All Layer mix run



           製程別
Case Study - Productivity Improvement

 • QIT 成員:
   – 製程單位
     • 掌握製程配方,了解製程限制
     • 提供可行之製程改善建議
   – 傳輸單位
     • 掌握傳輸資源,有效分配
     • 傳輸性能調校
   – 製程整合單位
     • 監控調校後的製程參數對產品的影響
   – 生產資訊系統單位
     • 調整自動派工規則
     • Right time to right tool
Case Study - Productivity Improvement

                             190.0

                                              TASUSG02 ChAB SELOX THK vs Time Chart
          idle 20mins        188.0


                             186.0
                                         Chamber A1
                                         Chamber A2
                             184.0       Chamber B1
                                         Chamber B2
                             182.0




                        nm
                             180.0


                             178.0


                             176.0


                             174.0


                             172.0


                             170.0
                                     0   5     10     15   20     25   30   35   40   45
                                                            time(s)
PDCA
                                    ◆ 改善機台: XXX
                                    ◆ 動作: 1. Chamber idle time extend to 20mins
◆ 提高 Chamber UT 與 POH 達成 T80 所            2. MES派工Rule修訂
需 capacity                                3. PPID Group 設定
                                          4. Dedicate OHT
                                    ◆ Owner : :製程 / IE / 傳輸 / MES




                                    ◆ 確認LQC data ok and no impact Cpk;
                                    check 2Lot
◆套用到機台, 持續觀察                        ◆ 確認新派工Rule的影響
◆ Owner : 製程 / 整合 / 傳輸 / IE / MES   ◆ 確認Dedicated OHT效能表現
                                    ◆ Owner : 整合 / 傳輸 / IE / MES
效益評估
一.有形效益:                           T90                            T80
                     K                              K
                    7                   6.5K   17                  16.5K
 1.在產能上的改善 :       6.5
                    6
                         5.5K                  16       15.5K
                   5.5                         15
                    5
                                               14
                   4.5            Gain 1K                        Gain 1K
                    4                          13
 2. 在節省成本方面:             Before     After               Before     After


 可滿足未來 T80 16.5K 的產能.不需再買新機台           省下購置新機台的成本
二.無形效益:
 1.藉由QIT活動與各部門合作, 讓同仁彼此更熟悉與了解對方的工作內容
 2.不需再新增機台,減少了製程與設備在裝機上的風險與maintain Loading
Summary
• 知識在網路上 ? No!!
 – 知識已在你我手上,已在組織內,關鍵在如何
   找出有價值的知識化為能獲利的策略和行動.
 – 光是記錄與呈現生產資訊是不夠的,更積極的要
   發掘隱含在內的知識,所以提出i-MES
• 知識的定價?
 – 在於對組織產生多少貢獻,例如本篇文章提出之
   良率的提昇,Cycle Time的縮短就是一例
 – 利用知識分享與團隊合作,制定策略並化為具
   體行動,透過PDCA循環,不斷改善,知識的
   定價可能就是所省下之購置機台的成本
Summary
• 如何成為專家? 留一手的問題?
 – 能找出隱含的知識的
 – 個別都不是專家,也無法從網路上、圖書館找
   到
 – 合起來看,這一群人(QIT)就是專家,面對公司
   共同的目標,不會也不應留一手
• 知識管理可以降低成本,增加獲利
 – Yield / Cycle Time improve
 – 提高Equipment Utilization (藉由QIT) , 節省新
   機台的購置成本
Q&A

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:12
posted:6/27/2012
language:
pages:26