You are on page 1of 52

Statistical

6

SUPPLEMENT
Process
Control

PowerPoint presentation to accompany
Heizer and Render
Operations Management, Eleventh Edition
Principles of Operations Management, Ninth Edition

PowerPoint slides by Jeff Heyl

© 2014 Pearson Education, Inc.

Learning Objectives
1. Apply quality management tools for
problem solving
2. Identify the importance of data in
quality management

6S–2
6S–2

Introduction
Statistical Process Control (SPC)

Statistical
Quality Control
Acceptance Sampling (AS)

 Statistical process control is a statistical technique that is
widely used to ensure that the process meets standards.
 Acceptance sampling is used to determine acceptance or
rejection of material evaluated by a sample.

6S–3
6S–3

Introduction
Pottery Making Process

Preparing
the clay for Wedging Throwing Pinching Painting Firing
throwing pots

6S–4
6S–4

Introduction

6S–5
6S–5

Natural Variation Variation Assignable Variation  Natural variation – can not be eliminated  Assignable variation -.Statistical Process Control Chart (SPC)  Variability is inherent in every process.Deviation that can be traced to a specific reason: machine vibration. tool wear. new worker. 6S–6 6S–6 .

6S–7 6S–7 . detect. Statistical Process Control Chart (SPC)  The essence of SPC is the application of statistical techniques to prevent. and eliminate defective products or services by identifying assignable variation.

Statistical Process Control Chart (SPC) A control chart is a time-ordered plot obtained from an ongoing process Abnormal variation Out of due to assignable sources control UCL Mean Natural variation due to chance LCL Abnormal variation due to assignable sources 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Sample number 6S–8 6S–8 .

good/bad. go/no go. Statistical Process Control Chart (SPC) x -charts (for controlling central tendency) Control Charts for Variable Data R-charts (for controlling variation) Control Charts p-charts (for controlling percent defective) Control Charts for Attribute Data c-charts (for controlling number of defects)  Variable Data (continuous): quantifiable conditions along a scale. such as speed.  Attribute Data (discrete): qualitative characteristic or condition. etc. such as pass/fail. 6S–9 6S–9 . length. density.

then the process is assumed to be out of control and corrective actions need to be taken. LCL and the measured values 4. 5. then the process is assumed to be in control and no actions should be taken except continuing to monitor. If all the measured values fall within the LCL and the UCL. 6S-10 . If one or more data points fall outside the control limits. Plot UCL.Statistical Process Control Chart (SPC) 1. Take random samples 2. Calculate the upper control limit (UCL) and the lower control limit (LCL) 3.

x-Charts Upper control limit (UCL) = x + A2R Lower control limit (LCL) = x .A2R where R = average range of the samples A2 = control chart factor x = average of the sample means 6S–11 6S–11 .

x-Charts Hour 1 Hour 2 Box Weight of Box Weight of Number Oat Flakes Number Oat Flakes 1 17 1 14 2 13 2 16 3 16 3 15 4 18 4 14 5 17 5 17 6 16 6 15 7 15 7 15 8 17 8 14 9 16 9 17 Range=18-13=5 Range=17-14=3 R = (5+3)/2 = 4 6S–12 6S–12 .

x-Charts Upper control limit (UCL) = x + A2R Lower control limit (LCL) = x .A2R where R = average range of the samples A2 = control chart factor from Table S6.1 (page241) x = average of the sample means 6S–13 6S–13 .

11+15.11 Average=(14+16+…+17)/9=15.22 x = (16.22)/2 = 15. x-Charts Hour 1 Hour 2 Box Weight of Box Weight of Number Oat Flakes Number Oat Flakes 1 17 1 14 2 13 2 16 3 16 3 15 4 18 4 14 5 17 5 17 6 16 6 15 7 15 7 15 8 17 8 14 9 16 9 17 Average=(17+13+…+16)/9=16.665 6S–14 6S–14 .

x-Charts Upper control limit (UCL) = x + A2R Lower control limit (LCL) = x .A2R where R = average range of the samples A2 = control chart factor x = average of the sample means 6S–15 6S–15 .

29 1.37 1.78 0.31 1.73 2.88 3.22 11 .26 6S–16 6S–16 .12 0 6 .x-Charts Sample Size Mean Factor Upper Range Lower Range n A2 D4 D3 2 1.00 0 7 .18 10 .28 0 5 .58 2.02 2.48 2.86 0.14 9 .92 0.34 1.42 1.74 0.58 0 4 .82 0.27 0 3 1.08 8 .

x-Charts Upper control limit (UCL) = x + A2R Lower control limit (LCL) = x .A2R where R = average range of the samples A2 = control chart factor from x = average of the sample means 6S–17 6S–17 .

Construct the x-chart with 3- control limit. x-Charts Example 1: Eight samples of seven tubes were taken at random intervals. Is the current process under statistical control? Why or why not? Should any actions be taken? Sample size = n = 7 A2 = ? 6S–18 6S–18 .

37 1.58 0 4 .73 2.27 0 3 1.22 11 .82 0.x-Charts Sample Size Mean Factor Upper Range Lower Range n A2 D4 D3 2 1.48 2.00 0 7 .14 9 .18 10 .31 1.26 6S–19 6S–19 .86 0.34 1.58 2.29 1.12 0 6 .28 0 5 .92 0.08 8 .78 0.02 2.88 3.74 0.42 1.

x-Charts Example: Eight samples of seven tubes were taken at random intervals..34 x  6.42 6.42(0.  6.42(0.43 oz LCL  x  A2 R  6.36  0.29 oz 6S–20 6S–20 ..17 oz 8 UCL  x  A2 R  6.38  .  0..36  0.. Is the current process under statistical control? Why or why not? Should any actions be taken? A2 = 0.16  0.36  6.36 oz 8 0. Construct the x-chart with 3- control limit.17)  6.18 R  0.18  .17)  6.

x-Charts Control Chart for sample of 7 tubes 6. 6S–21 6S–21 . No actions need to be taken except to continuously monitor this process.43 = UCL 6.36 = Mean 6.29 = LCL | | | | | | | | | | | | 1 2 3 4 5 6 7 8 9 10 11 12 Sample number It is assumed that the central tendency of process is in control with 99.73% confidence.

Use the sample statistic to calculate control limits 3. Plot control limits and measured values 4.Steps in Creating Charts 1. Investigate possible assignable causes and take actions 6S–22 6S–22 . Determine the state of the process (in or out of control) 5. Take samples from the population and compute the appropriate sample statistic 2.

R-Charts Upper control limit (UCL) = D4R Lower control limit (LCL) = D3R where R = average range of the samples D3 and D4 = control chart factors 6S–23 6S–23 .

18 10 .86 0.88 3.74 0.58 2.58 0 4 .22 11 .12 0 6 .00 0 7 .27 0 3 1.42 1.73 2.14 9 .31 1.R-Charts Sample Size Mean Factor Upper Range Lower Range n A2 D4 D3 2 1.29 1.02 2.28 0 5 .82 0.37 1.78 0.08 8 .34 1.48 2.26 6S–24 6S–24 .92 0.

R-Charts Example 2 Average range R = 5.3 pounds Sample size n = 5 From Table 1 D4 = ? D3 = ? 6S–25 6S–25 .

78 0.48 2.92 0.82 0.58 0 4 .28 0 5 .R-Charts Sample Size Mean Factor Upper Range Lower Range n A2 D4 D3 2 1.29 1.18 10 .27 0 3 1.88 3.12 0 6 .22 11 .42 1.74 0.02 2.08 8 .86 0.26 6S–26 6S–26 .73 2.37 1.34 1.31 1.14 9 .58 2.00 0 7 .

1 D4 = 2.12)(5.3) = 0 pounds 6S–27 6S–27 .3 pounds Sample size n = 5 From Table S6.12.3) = 11.2 pounds Mean = 5.3 LCLR = D3 R LCL = 0 = (0)(5. R-Charts Example 2 Average range R = 5.2 = (2. D3 = 0 UCLR = D4R UCL = 11.

R-Charts Example 3: Refer to Example 1... Eight samples of seven tubes were taken at random intervals.  0.17 oz 8 n=7 D3 =? D4 = ? 6S–28 6S–28 . Construct the R-chart with 3- control limits. Is the current process under statistical control? Why or why not? Should any actions be taken? 0.18 R  0.18  .16  0.

48 2.78 0.27 0 3 1.00 0 7 .37 1.29 1.18 10 .42 1.73 2.28 0 5 .82 0.08 8 .88 3.02 2.86 0.74 0.12 0 6 .22 11 .R-Charts Sample Size Mean Factor Upper Range Lower Range n A2 D4 D3 2 1.26 6S–29 6S–29 .92 0.14 9 .31 1.34 1.58 0 4 .58 2.

16  0.08(0.08.18  .18 R  0.17)  0.  0. R-Charts Example S6.17 oz 8 D3 =0.3: Refer to Example S6..01 oz 6S–30 6S–30 .92(0.92 UCL  D4 R  1.1.17)  0.. Is the current process under statistical control? Why or why not? Should any actions be taken? 0. D4 = 1.33 oz LCL  D3 R  0. Eight samples of seven tubes were taken at random intervals. Construct the R-chart with 3- control limits.

R-Charts Control Chart for sample of 7 tubes 0.01 = LCL | | | | | | | | | | | | 1 2 3 4 5 6 7 8 9 10 11 12 Sample number The variation of process is in control with 99.73% confidence. 6S–31 6S–31 .33 = UCL 0.17 = R 0.

but its variation is not in control. Mean and Range Charts (a) The central tendency of process is in control. UCL (R-chart detects R-chart increase in dispersion) LCL UCL (x-chart does not x-chart detect dispersion) LCL 6S–32 6S–32 .

but its central tendency is not in control.Mean and Range Charts (b) The variation of process is in control. UCL (R-chart does not R-chart detect changes in mean) LCL UCL (x-chart detects x-chart shift in central tendency) LCL 6S–33 6S–33 .

0)  6. and D4 = 2. Is the entire process under statistical control? Why or why not? n=4 D3 = 0.4: Seven random samples of four resistors each are taken to establish the quality standards.28(3. R-Chart and X-Chart Example S6.0 UCL  D4 R  2.28 R = (3 + 2 + … + 4)/7 = 3.0)  0 6S–34 6S–34 . Develop the R-chart and the x-chart both with 3- control limits for the production process.84 LCL  D3 R  0(3.

73% confidence. 6S–35 6S–35 .84 = UCL 3. R-Chart and X-Chart Control Chart for sample of 4 resistors 6.0 = R 0 = LCL | | | | | | | | | | | | 1 2 3 4 5 6 7 8 9 10 11 12 Sample number The variation of process is in control with 99.

79  0.73(3.5 + 101.0)  101.6 6S–36 6S–36 .0)/7  99.98 LCL  x  A2 R  99.0)  97.5 + … + 101.79  0.79 UCL  x  A2 R  99.73(3.R-Chart and X-Chart n = 4. A2 = 0.0 X= (100.73 R = (3 + 2 + … + 4)/7 = 3.

R-Chart and X-Charts Control Chart 101. 6S–37 6S–37 .7% confidence. with 99.73% confidence.6 = LCL | | | | | | | | | | | | 1 2 3 4 5 6 7 8 9 10 11 12 Sample number The central tendency of process is not in control with 99. In conclusion.79 = Mean 97.98 = UCL 99. the entire resistor production process is not in control since its central tendency is out of control although its variation is under control.

02 4 3.01 0. Calculate the control limits.00 0.05 10 4. plot the control charts.06 6 3.04 9 3.01 0.02 0.02 7 4.97 0.99 0.98 0.02 8 3. and determine if the process is in control No.06 6S–38 6S–38 .Problem A part that connects two levels should have a distance between the two holes of 4”. The following ten samples of size four were collected.98 0.06 3 4.00 0.04 2 3.99 0. of Sample Mean Range 1 4. It has been determined that x-bar chart and R-chart should be set up to determine if the process is in statistical control.05 5 4.

5: Resistors for electronic circuits are manufactured at Omega Corporation in Denton. Develop the R-chart and the chart both with 3- control limits for the production process. R-Chart and X-Chart Example S6. The head of the firm’s Continuous Improvement Division is concerned about the product quality and sets up production line checks. TX. Is the entire process under statistical control? Why or why not? # of sample Readings of Resistance (ohms) 1 99 100 102 101 2 101 103 101 101 3 98 102 101 99 4 99 100 99 100 5 99 99 98 100 6 95 100 97 96 7 101 99 101 103 6S–39 6S–39 . She takes seven random samples of four resistors each to establish the quality standards.

0 = LCL | | | | | | | | | | | | 1 2 3 4 5 6 7 8 9 10 11 12 Sample number 6S–40 6S–40 .  4 n=4 D3 =0 D4 = 2..0 97..R-Chart and X-Chart # of Sample 1 2 3 4 5 6 7 Sample range 3 2 4 1 2 5 4 Sample mean 100.0 99.0 = R is in control with 99.84 6.28  3.5 100.28 R  3.0 101.84 = UCL LCL  D3 R  0 variation of process 3.0  6.5 99.0 7 UCL  D4 R  2.73% confidence.0 3  2  .5 101.

6 = LCL Thus. 1 2 3 4 5 6 7 8 9 10 11 12 Sample number 6S–41 6S–41 .0)  102.5 + … + 101. 97.73(3.0 = UCL central tendency of process is not in control 99.0 n=4 LCL  x  A2 R  99.5 99..5 100.0)  97. entire process is | | | | | | | | | | | | not in control.. R-Chart and X-Chart # of Sample 1 2 3 4 5 6 7 Sample range 3 2 4 1 2 5 4 Sample mean 100.8  0.0)/7  99.0 99.73 UCL  x  A2 R  99.8 = X with 99.8 R  3.6 102.73(3.0 7 A2 =0.5 101.  4 X= (100.73% confidence.8  0.0 97.0 3  2  .0 101.

Each day last week. Set up control charts to determine if the process is in statistical control Day Package 1 Package 2 Package 3 Package 4 Monday 23 22 23 24 Tuesday 23 21 19 21 Wednesday 20 19 20 21 Thursday 18 19 20 19 Friday 18 20 22 20 6S–42 6S–42 .EX 2 in class A quality analyst wants to construct a sample mean chart for controlling a packaging process. he randomly selected four packages and weighed each. The data from that activity appears below.

density. length. Statistical Process Control Chart (SPC) x -charts (for controlling central tendency) Control Charts for Variable Data R-charts (for controlling variation) Control Charts p-charts (for controlling percent defective) Control Charts for Attribute Data c-charts (for controlling number of defects)  Variable Data (continuous): quantifiable conditions along a scale.  Attribute Data (discrete): qualitative characteristic or condition. go/no go. etc. such as pass/fail. 6S–43 6S–43 . such as speed. good/bad.

Control Charts for Attribute Data  Categorical variables  Good/bad. acceptable/unacceptable  Measurement is typically counting defectives  Charts may measure  Percentage of defects (p-chart)  Number of defects (c-chart) 6S–44 6S–44 . yes/no.

P-Charts p (1  p ) UCL  p  z pˆ  p  z n p (1  p ) LCL  p  z pˆ  p  z n where p = mean percent defective overall the samples z = number of standard deviations = 3 n = sample size 6S–45 6S–45 .

01 3 0 .03 9 3 .03 19 0 .02 20 4 . P-Charts Example S6. Develop a p-chart with 3- control limits and determine if the process is in control.05 6 2 .04 15 5 .11 8 3 .01 14 7 .04 Total = 80 6S–46 6S–46 .05 12 1 .07 5 4 . Sample Number Percent Sample Number Percent Number of Errors Defective Number of Errors Defective 1 6 .03 18 3 .05 17 11 .06 2 5 .04 7 5 .06 11 6 .6: Data-entry clerks at ARCO key in thousands of insurance records each day.00 10 2 .00 13 8 .02 16 4 .08 4 1 . One hundred records entered by each clerk were carefully examined and the number of errors counted.

P-Charts n = 100 Total number of errors 80 p   0.  0.04) LCL  p  z  0.04(1  0. p   0..099 n 100 p (1  p ) 0.019  0 n 100 Because we cannot have a negative percent defective 6S–47 6S–47 .04  3   0.04 or ..04  3   0.04) UCL  p  z  0.040 Total number of records examined (100)(20) Total fraction defective 0.04(1  0.040 Number of samples 20 p (1  p ) 0.06  0.05  .

02 – .04 – p = 0. 6S–48 6S–48 .00 .10 – UCL= 0. P-Charts Possible assignable causes present .11 – .10 .00 – 2 4 6 8 10 12 14 16 18 20 Sample number Possible good assignable causes present The process is not in control with 99.07 – .03 – .01 – | | | | | | | | | | LCL= 0.09 – Percent defective .08 – .05 – .06 – .73% confidence.04 .

C-Charts  A c-chart is used when the quality cannot be measured as a percentage.  Number of car accidents per month at a particular intersection  Number of complaints the service center of a hotel receives per week  Number of scratches on a nameplate  Number of dimples found on a metal sheet 6S–49 6S–49 .

3 c where c = mean number defective overall th 6S–50 6S–50 .C-Charts UCL = c + 3 c LCL = c .

6. Determine the 3.35 => 0 The process is in control Because we cannot have the negative with 99.7: Over 9 weeks. C-Charts Example S6. 4.73% confidence. 8. 7.35 4 – 2 – LCL = 0 LCL = c . 9. 8.3 c 0 – | | | | | | | | | =6-3 6 1 2 3 4 5 6 7 Week Number 8 9 = .35 c = 54 / 9 = 6 complaints /week 14 – Number of defect 12 – 10 – UCL = c + 3 c 8 – =6+3 6 6 – c= 6 = 13. number of defective records 6S–51 6S–51 . control limits of a c-Chart. UCL = 13.1. 0. Red Top Cab company received the following numbers of calls from irate passengers: 3. for a total of 54 complaints. 9.

Managing Quality Summary 1. There are multiple tools to identify the relationships between variables 6S–52 6S–52 . Effective quality management is data driven 2. There are multiple tools to identify and prioritize process problems 3.