SUNON Nan Hai PlantAttribute Measurement Systems計數型測量系統SUNON Nan Hai PlantAttribute Terminology 計數術語計數術語qAttribute Data:計數型數據 Qualitative(go/no go) data that can be counted for recording and analysis 可用作記錄或分析的定性數據(通常為合格/不合格) qAttribute Measurement System(AMS):計數型測量系統 A measurement system that compares a product to a standard and accepts the product if in meets the standard . 將每一產品與標準對比,如果符合標準的要求則接受 Can be inspected by an inspection or by using a go/no go gauge 可由檢驗員或合格/不合格量規來實施檢查qScreen:篩選 100% inspection using an AMS 用計數測量系統百分百評價產品SUNON Nan Hai Plant Attribute Terminology 計數術語計數術語q Screen Effectiveness:篩選效率 The ability of the AMS to discriminate good from bad AMS 區分合格與不合格產品的能力q Customer Bias : 客戶偏好 Screening criteria is too tight:Good product is being rejected 篩選標準太嚴格:合格產品被拒收q Producer Bias:生產者偏好 Screening criteria is too loose:Defective product is being accepted 篩選標準太寬鬆:不合格產品被接收SUNON Nan Hai PlantThe Purpose of Attribute R & R計數值計數值R & RR & R的目的的目的q To understand if workers across all shifts,machines and lines use the same criteria to determine “pass” or “fail” 確定各班次、各機器、各流水線的檢驗員是否用同一標準區分 合格與不合格.q To quantify the ability of inspectors or gauges to repeat their inspection results and conclusions 量化測量者或測量儀器是否能正確重复檢驗結果的能力SUNON Nan Hai PlantThe Purpose of Attribute R & R計數值計數值R & RR & R的目的的目的qTo understand the level of conformance to a standard 確定檢驗員/量儀符合標準的程度,包括: The frequency where inspectors ship truly defective product (Producer Bias) 檢驗員接受不合格產品的概率 The frequency where inspectors do not ship truly acceptable Product (Customer Bias) 檢驗員拒收合格品產品的概率SUNON Nan Hai PlantThe Purpose of Attribute R & R計數值計數值R & RR & R的目的的目的qIdentify the following:識別以下各項 Training need (Individual(s) or for all) 培訓需求(個別或全部檢驗員) The lack of documented procedures or control plans 缺少程序或控制計劃 Whether standards are clearly defined 標準沒有清晰定明 Adjustment or correlation of gauges 量儀需要調整或進行相關對比SUNON Nan Hai PlantPoints for Attribute Measurement Systems計數測量系統應注意事項計數測量系統應注意事項1.Attribute data is normally generated from screening of product 計數值通常在篩選產品中產生1.100% product screening is conducted when the process is incapable of meeting the required acceptable level 100%產品篩選通常是在過程不能以高比例生產合格產品的 時,實施1.As a result,many borderline products(both good and defective) may occur 因此,將生產大量的邊界產品(合格或不合格產品)SUNON Nan Hai PlantPoints of Attention for Attribute Measurement Systems計數測量系統應注意事項計數測量系統應注意事項qVariation exists in any measurement system,inspection and operational process.This will lead to the acceptance of truly defective products or rejection of truly acceptable products. 很多測量系統,檢驗和運作過程都存在變異,這導致大量合格產品的 拒收或大量不合格產品的接受As a result,we should Focus on improving theProcess capability and Not on improving the Screening systems因此我們應該將重點放在改進系統能力而非完善篩選系統 SUNON Nan Hai PlantThe Methods for Conducting Attribute R & R計數計數 R & R方法方法1)Select a minimum of 30 products from the process 從過程中選取至少30件產品 There will be products in the study where they are 在我們選出做分析的產品中應包括下列產品: Defective 不良品 Defect free 無缺陷產品 Marginal cases 邊緣產品SUNON Nan Hai PlantThe Methods for Conducting Attribute R & R計數計數 R & R方法方法2) Select the inspectors (who should be qualified and experienced) and the gauges 挑選檢驗員和量儀.檢驗員應具有經驗和資格3) Let the inspectors,independently and in random order,inspect the products and determine whether they pass or fail. Repeat this process 讓每位檢驗員獨立以隨機的次序檢驗這些產品並確定合格與否, 重复.SUNON Nan Hai PlantThe Methods for Conducting Attribute R & R計數計數 R & R方法方法4) Enter the data into the Attribute R & R.Excel spreadsheet or Minitab to calculate the effectiveness of the AMS 將數據輸入至Attribute R & R. Excels工作表或Minitab 報告MSA的檔案5) Record the results and review the process. 記錄結果.如果要采取恰當的措施修正檢驗過程6) Prepare the study to verify the action taken. 重新進行分析,以驗證修正措施SUNON Nan Hai PlantAttribute R & R-Worksheet計算計算R&R-方法方法1利用利用Excel表表 Appraiser Score=檢查員自身一致次數總次數Score vs.Attribute=檢查員檢測結果与已知標準相符的次數總次數Screen Effective score=所有檢查員有一致性的次數總數次Screen Effective Score VS. Attribute=所有檢查員與標準相一致次數總次數SUNON Nan Hai PlantDefinition of Terminology for Attribute R & RExcel Spreadsheet 例題解析例題解析q Appraiser Score in the consistency within one person. 檢驗員的評分(% Appraiser Score)評價的是個人的一致性 （重复性）q score vs. Attribute is a measure of how well the operators evaluation agrees with that of the “expert”. 檢驗員評分比較標準(%Score vs. Attribute)評價的是檢驗員和“專家”評價相符的程度qScreen % Effective Score is a measure of how well the inspectors agree with each other. 檢驗員之間的有效評分(Screen % Appraiser Score)評價的是檢驗員之間的一致性（再現性）qScreen % Effective Score VS. Attribute in an overall measure of consistency between inspectors and agreement with “expert”. 檢驗員之間有效評分再比較標準(Screen % Effective Score vs. Attribute)綜合評價全部檢驗員和專家的一致性 越接近100%越好SUNON Nan Hai PlantGuidelines for Selecting MSA Samples MSA樣本指南樣本指南qUsually about 30 samples,2-4inspectors (if several inspectors use the gauge),2-3 trials 通常要大概30個樣本，2-4個檢驗員(如果多個檢驗員使用量儀),2-3次檢驗qDepending on the purpose of the study 1 or more gauges will be used 根據研究的目的，可用個或多個量儀qSelect enough samples such that number of samples x number of inspectors/gages15 一般來說，選擇足夠的樣本使樣本數目檢驗員數目量儀數目15SAMPLE SELECTION 樣本選擇Option 1: if process variability is unknown,the samples selected should cover the full range of the normal process/product variation (go get TV) 選擇1:如果過程變異未知，樣本應包含正常過程/產品變異的全部範圍 (以獲得TV) Option 2:if process variability is Known,the samples selected should uniformly span beyond the width of the specifications. 選擇2:如果過程變異已知，樣本應在規格範圍內平均分布SUNON Nan Hai PlantAttribute R & R-Using Minitab 計量計量R&R-方法方法2利用利用Minitab打開：“Attribute MSA.mtw”SUNON Nan Hai PlantAttribute R & R-Using Minitab 計量計量R&R-方法方法2利用利用Minitab填寫完填寫完后后,點點“ok”喔喔!SUNON Nan Hai PlantAttribute R & R-Using Minitab 計量計量R&R-方法方法2利用利用MinitabSUNON Nan Hai PlantAttribute R & R-Using Minitab 計量計量R&R-方法方法2利用利用MinitabWithin Appraisers Within Appraisers Assessment AgreementAppraiser # Inspected # Matched Percent 95 % CI1 14 14 100.00 (80.74, 100.00)2 14 11 78.57 (49.20, 95.34)3 14 14 100.00 (80.74, 100.00)# Matched: Appraiser agrees with him/herself across trials.Each Appraiser vs. Standard Each Appraiser vs. Standard Assessment AgreementAppraiser # Inspected # Matched Percent 95 % CI1 14 11 78.57 (49.20, 95.34)2 14 9 64.29 (35.14, 87.24)3 14 10 71.43 (41.90, 91.61)# Matched: Appraisers assessment across trials agrees with theSUNON Nan Hai PlantAttribute R & R-Using Minitab 計量計量R&R-方法方法2利用利用MinitabAssessment DisagreementAppraiser # pass / fail Percent # fail / pass Percent # Mixed Percent1 2 28.57 1 14.29 0 0.002 1 14.29 1 14.29 3 21.433 1 14.29 3 42.86 0 0.00# pass / fail: Assessments across trials = pass / standard = fail.# fail / pass: Assessments across trials = fail / standard = pass.# Mixed: Assessments across trials are not identical.Between Appraisers Between Appraisers Assessment Agreement# Inspected # Matched Percent 95 % CI 14 8 57.14 (28.86, 82.34)# Matched: All appraisers assessments agree with each other.All AppraisersAll Appraisers vs vs Standard Standard Assessment Agreement# Inspected # Matched Percent 95 % CI 14 6 42.86 (17.66, 71.14)Lower bound must be80%SUNON Nan Hai PlantSUNON Nan Hai PlantSUNON Nan Hai Plant