1、1FORD 8D Methodology / DMAIC Approach & PDCA.2PDCA introduction8D introduction DMAIC brief introductionCIT introduction.3PDCAPDCA Cycle: Plan 方案 Do 執(zhí)行 Check 檢查 Action 糾正活動.4PDCA 改良方式C-Check檢查A-Action糾正活動D-Do執(zhí)行P-Plan方案Continuous ImprovementPDCA是一個繼續(xù)的質(zhì)量改善的螺旋上升循環(huán).5優(yōu)質(zhì)企業(yè)客戶稱心PDCAPDCAPDCAPDCA 繼續(xù)改良PDCA不是一個平

2、面環(huán)，它是階梯式上升的，這個循環(huán)走完一遍應(yīng)該就曾經(jīng)處理一部分問題，獲得一部分成果，到了循環(huán)的下一個周期，在范圍上就應(yīng)該產(chǎn)生了變化，所以說它不是一個平面環(huán)。.6Ford-8D PracticeFollow this template!.7運用 Ford 8D 的時機 Consequences identified因果關(guān)系的鑒別 Cause unknown緣由不明時用8D來找緣由 Resources for problem solution & recurrence prevention management-approved 處理問題的資源和防止再發(fā)生措施的同意 Single-person ca

3、pacity for problem-solving exceeded不是一個人能處理的，需求經(jīng)過腦力激蕩和團隊協(xié)作來處理 Symptoms defined異常情況曾經(jīng)出現(xiàn)，并很明顯 Symptom quantified with measurement異常情況可量化.8Ford-8D WorkflowDescribe the ProblemD1: 主題選定及小組成立D2: 問題描畫及緣由確認(rèn)D3: 展開暫時防堵措施D4: 定義及驗證真因D5: 列出, 選定及驗證永久對策D6: 執(zhí)行永久對策及效果確認(rèn)D7: 預(yù)防再發(fā)及規(guī)范化D8: 恭賀團隊及未來方向Prevent RecurrenceImpl

4、ement Permanent Corrective ActionRecognize Team & Individual ContributionsYNEstablish a TeamDevelop Interim Containment ActionDefine & Verify Root CauseChoose & Verify Permanent Corrective Actions.9D1: Team BuildingEstablish a team with people:與此project 或任務(wù)相關(guān)的人有cross function的人一同參與有相關(guān)的knowledge expe

5、rience定期每周/每月review進度.10D2: Problem Description What is wrong with what? How can the problem be quantified? What? When? Where? How much? Cause & Effect? Something changed / never been successful?.11D3: Development of ICAICA = Interim Containment Action (暫時防堵措施) Quick fixes Temporary detective action

6、s Isolation from harmful effects. Short-term action before PCA are taken.PCA = Permanent Corrective Action.12D4: Definition & Verification of Root Cause Identify “real problem- 5 why Test each possible cause. Systematic use of resources. Ideality (What it is in best way)NO?YES?.13 Root Cause Identif

7、ication - SamplingProblem: Bipolar transistor gain in BiCMOS process out-of-spec high at e-test (WAT). 在WAT測試中發(fā)現(xiàn)BiCMOS工藝中的雙極晶體管增益高出規(guī) 范的要求 Why#1 Why was the bipolar transistor gain too high? 為什么這個雙極晶體管的增益太高？Answer#1: Emitter junction was too deep. 發(fā)謝極太深How do you know when you have found the true roo

8、t cause? there must be some element of human decision. .14Why#2: Why was the emitter junction too deep? 為什么發(fā)謝極太深？Answer#2: Actual furnace temperature was slightly high. (confirmed by manual profiling) 爐管的實踐溫度偏高了一點Why#3: Why was the actual furnace temperature slightly high? 為什么實踐的溫度會偏高一點？Answer#3: Fu

9、rnace controller spike thermocouples (T/Cs) did not extend far enough into the heat zone. 爐管的熱電偶控制器不可以控制到爐管加熱的整個區(qū)域 Root Cause Identification - Sampling.15 Root Cause Identification - SamplingWhy#4: Why didnt the spike thermocouples not extend far enough into the heat zone? 為什么爐管的熱電偶控制器不可以控制到爐管加熱的整個區(qū)

10、 域？Answer#4a: Thermocouples were slightly shorter than previous ones. 這個熱電偶比之前的一個短Answer#4b: T/C position was not checked after installation. 安裝后熱電偶的位置沒有檢查熱電偶的位置偏移了 .16D4: Definition & Verification of Root CauseWhy#5a: Why were the spike T/Cs slightly shorter than previous ones? 為什么這個熱電偶比之前的一個短？Answ

11、er#5a: Spike T/Cs were being purchased from a new vendor. 這個熱電偶是從一個新的供應(yīng)商那里采購的 Why#6a: Why was a new spike T/C vendor being used. 為什么要從這個新的供應(yīng)商處采購？Answer#6a: Reduced costs. 降低本錢 Why#7a: Why wasnt this problem anticipated? 為什么沒有預(yù)料到這個問題？Answer #7a: No guidelines or procedures for such changes (no physic

12、al comparison done and no “What-could-go-wrong? analysis). 對這樣的變卦沒有程序規(guī)定沒有做對比實驗，也沒有“能夠 會引起什么樣的錯誤的分析“沒有程序文件或檢查表是一個經(jīng)常導(dǎo)致問題發(fā)生的根本緣由。.17D4: Definition & Verification of Root CauseWhy#5b: Why was the T/C position not checked after installation? 為什么安裝后熱電偶的位置沒有檢查Answer#5b: Equipment tech did not expect any pr

13、oblem. 設(shè)備偵探?jīng)]有預(yù)期有任何問題。 Why#6b: Why did the Equip Tech not expect any problem. 為什么設(shè)備偵探?jīng)]有預(yù)期有任何問題？Answer#6b: He was not informed that a new spike T/C vendor was being used. 沒有通知他運用了新廠商的熱電偶 Why#7b: Why was Equip Tech not informed that a new spike T/C vendor was being used? 為什么沒有通知他運用了新廠商的熱電偶Answer #7b: N

14、o procedure to do so for such material changes. 對于這個資料的變卦，沒有程序規(guī)定要這樣做?！皼]有程序文件或檢查表是一個經(jīng)常導(dǎo)致問題發(fā)生的根本緣由。.18下面列舉了一些在FAB反復(fù)發(fā)生的典型的根本緣由 沒有程序、或程序不完善、或 程序沒有文件化或沒有檢查表 文件化的程序不清楚、不明確、不完善或讓人糊涂。 騰躍式的程序以節(jié)省時間.D4: Definition & Verification of Root Cause.19D5: Choice & Verification of PCAPCA = Permanent Corrective Action

15、Elimination of root cause? Costs? Production stop? Overtime work necessary for implementation? Overtime work necessary for maintenance? .20D5: Choice & Verification of PCA針對每一“為什么的答案都應(yīng)至少有一個措施這也就意味著一個充分的預(yù)防處理方案至少有57個措施一種常犯的錯誤是把偵測措施作為預(yù)防措施。如檢驗、丈量、檢查、監(jiān)控、做先行批等。偵測措施是必需的，但不是充分的它們只是遏制了這個問題。偵測措施與消費沒有直接的關(guān)系它們不能

16、對產(chǎn)品增值只是添加了本錢。它們就象買“保險，保險是有本錢的保險費.21D6: Implementation & Validation of PCA ICA removed Long-term result monitored.22D7: Prevention of Recurrence System modification according to PCA Improvement standardizationFollow me!文件化的程序并進展適當(dāng)?shù)呐嘤?xùn)才是真正的預(yù)防措施 .23預(yù)防措施可采取以下的格式進展：建立或修正程序針對沒有程序或不完善程序程序的文件化(建立/修正文件)培訓(xùn)文件的執(zhí)

17、行消除環(huán)境要素 D7: Prevention of Recurrence.24D8: Recognition of Team & Individual ContributionsAppreciation for every members effort !.25DEFINE the customer, their CTQ (Critical to Quality) issues, and the core process involved.MEASURE the performance of the core process involved.Stage 1: 定義改善時機Stage 2: 丈

18、量運作程度Stage 3: 分析改善時機Stage 4: 改善運作程度Stage 5: 控制運作程度CONTROL the Improvements to keep the process on the new course.IMPROVE the target process by designing creative solutions to fix and prevent problems.Changeoperations?Re-define the operationsNoYesANALYZE the data collected to determine root cause of

19、defects and opportunities for improvement.DMAIC Workflow.26Stage 1: Define Define who customers are, what their requirements are for products and services, and what their expectations are Define project boundaries? The stop and start of the process Define the process to be improved by mapping the process flow.27Stage 2: Measure Develop a data collection plan for the process Collect data from many sources to determine types of defects and m