08 CM0471 Module 8

8/22/2019 08 CM0471 Module 8

http://slidepdf.com/reader/full/08-cm0471-module-8 1/37

Six Sigma Black Belt

Cert. Prep. Course:ControlControlControlControlModule VIII

©2009 ASQ 2

Agenda

This module consists of four lessons:

1. Statistical process control (SPC)

2. Other control tools3. Maintain controls4. Sustain improvements

©2009 ASQ 3

Lesson 1 – Statistical ProcessControl (SPC)

Define and describe the objectives of SPC, includingmonitoring and controlling process performance,

tracking trends, runs, etc., and reducing variation

in a process. (Understand)

VIII.A.1 Statistical Process Control

8/22/2019 08 CM0471 Module 8


©2009 ASQ 4

Statistical Process Control (SPC)

Definition• Using samples (rational subgroups), SPCestablishes the limits of natural variation for a

process.

Dr. Shewhart’s pioneering work in the area of

process analysis enabled generations of qualityengineers to adopt an effective approach to dataanalysis and process monitoring.

• SPC is used by organizations to measureprocesses (not just manufacturing).• It is a feedback system for the “operator” to identifythe need for process intervention.

©2009 ASQ 5

Benefits

• Assure customers that production is consistent over time

• Identify successful process improvement• Increase product consistency

• Maintain product quality

• Reduce the need for inspection

• Increase production yield

Objectives of SPC

• Produce data to inform and guide process improvement

• Reduce variation

• Increase knowledge about the process

• Detect, in real time, occurrences of special causes


©2009 ASQ 6

SPC tools achieve the objectives by collecting andanalyzing data

• Control limits are placed on the control chart to show

three standard deviations above and below averageor central line (these are NOT specification limits!)

• Subgroup data taken over time are plotted on the

control chart

• Subgroup statistics (mean/median and standard

deviation) are compared to the control limits to

determine process stability

• Therefore, the control chart is used to help youunderstand the variation in process


8/22/2019 08 CM0471 Module 8


©2009 ASQ 7

• Variation can be

– Common cause

– Special (assignable) cause

• Terminology

– “In statistical control” means the processis stable

– “Out of statistical control” means the processis unstable

• Predict the performance of a process using trendsand/or patterns

• Monitor the process as a part of the feedbacksystem directed at continuous improvement.


©2009 ASQ 8

• Common Cause

– Variation that is inherent to the system

– Management is responsible for reducing thisvariation

– Values that are located between the controllimits

– When a control chart contains points that arelocated between the control limits only, then

the process is said to be in statistical control(stable).

Common vs. Special Cause of Variation


©2009 ASQ 9

• Special Cause

– Variation that is not inherent to the system

– Employees are responsible for determining thesecauses

– Values are located outside the control limits

– Causes for which the reasons can be identified

– When a control chart contains some points thatfall outside of control limits, then the process issaid tonot be in statistical control (not stable).

Common vs. Special Cause of Variation

Basic rule of SPCVariation from common-cause systems should be left to chance,but special causes of variation should be identified and eliminated.


8/22/2019 08 CM0471 Module 8


©2009 ASQ 12

Progress Check

Which of the descriptions in the table below are

examples of Statistical Process Control? Check theappropriate box at the right.

Description/Name Example Non-Example

100% inspection (screening) of all production

Recording when assignable (special) causesare detected and taking corrective action

periodically (say, weekly)

Determining the natural variation of a process

Assigning design engineers to maintain SPC

Using samples to determine process stability

©2009 ASQ 14

Identify and select critical characteristics for controlchart monitoring. (Apply)


VIII.A.2 Selection of Variables

©2009 ASQ 15

Selection of Variables

Choosing the characteristic to be charted:• A variable for an SPC chart is selected for

monitoring purposes.

• It is usually the critical dimension / feature of aproduct / service being measured

• Give high priority to characteristics that are currently

running with a high defective rate

• Identify process variables

that contribute to end

product characteristics

SELECT

MEASUREMENT

MEASUREMENT

TYPE

VARIABLE ATTRIBUTE

8/22/2019 08 CM0471 Module 8


©2009 ASQ 16

Choosing the characteristic to be charted:

• Verify that the measurement process has sufficientaccuracy and precision to provide data (Gage R&R)

that does not obscure the variation in the process

• Determine the earliest point in the process at whichdata collection could be done, so that the SPC chart

serves as a warning device

Selection of Variables

SELECT

MEASUREMENT

MEASUREMENT

TYPE

VARIABLE ATTRIBUTE

©2009 ASQ 17

Progress Check

As a group, discuss some historical sourcesthat your organization uses for identifying keycharacteristics for monitoring purposes.

©2009 ASQ 18

Define and apply the principle of rational subgrouping.(Apply)


VIII.A.3 Rational Subgrouping

8/22/2019 08 CM0471 Module 8


©2009 ASQ 19

Rational Subgrouping

Donald J. Wheeler has six guiding principles for subgroupingi n a rational manner.

• Never knowingly subgroup unlike things together

• Minimize variation within each subgroup

• Maximize opportunity for variation between subgroups

• Average across noise, not across signals

• Treat charts in accordance with the use of the data

• Establish standard sampling procedures

Definition

•Rational subgroups are subgroups of data collected underrelatively homogeneous conditions.

•A rational subgroup is a subset of data defined by a specificfactor such as a stratifying factor or a time period.

•Rational subgrouping identifies and separates special causevariation (variation between subgroups) caused by specific,identifiable factors.

©2009 ASQ 21

Rational Subgrouping – Example

A quality engineer (QE) desires to monitor a process

that manufactures PET (plastic) bottles for thebeverage industry. The bottles are injection-moldedon a multi-cavity carousel. The particular carouselcontains four cavities, and the QE initially decides to

take three bottles from each cavity each hour and

measure a criticalcharacteristic. The

data might looklike the table here,where M1, M2,and M3 are thethree measures.

Therewouldbe aseparatecontrolchart for

eachcavity

©2009 ASQ 22

Progress Check

A candy-making process uses 40 pistons to deposit 40chocolate pieces in a 5 x 8 array on a moving sheet of waxpaper. Below are two options that illustrate how a rational

subgrouping of sample size n = 5 may be selected:Option 1: the first five chocolates in each row formed by five

different pistons.Option 2: the upper left-hand chocolate formed in five

consecutive arrays by the same piston.

Check off the optimal box.Use a multi-vari chart to determine

where the variation exists: piece-to-piece(Option 1) or within piece (Option 2)

8/22/2019 08 CM0471 Module 8


©2009 ASQ 23

Select and use the following control charts in varioussituations: X bar-R, X bar-s, individual and movingrange (ImR), p, np, c, u, short-run SPC and moving

average. (Apply)

VIII.A.4 Control Chart Selection


©2009 ASQ 24

Control Charts

• Originated by Walter Shewhart. Also known asShewhart Charts and Statistical Process ControlCharts.

• BenefitsDetection of special or assignable causes ofvariationIdentified by shifts in either location (mean) orspread (standard deviation)

Save money• Uses

Attain a state of Statistical ControlMonitor a processDetermine process capability

©2009 ASQ 25

Types of Control Charts• Attribute Charts• Variable Charts

8/22/2019 08 CM0471 Module 8


©2009 ASQ 26

X-bar and R (X-bar and Range)

X-bar estimates the process central tendency or location(mean) over time and the range (R) estimates the dispersionor process spread (standard deviation) over time.

Classification

Count

RK5

©2009 ASQ 27

Steps for Calculating X-bar and R

1. Collect data by subgroup

2. Calculate the mean for each group

3. Calculate the range for each subgroup

4. Estimate the mean of the population by calculatingthe grand mean.

The subgroup mean is the sum of the samples divided by

the number of samples:n = number of samples

An example has eight subgroups of sample size n = 5

The subgroup range is the maximum value minus the

minimum value:

R = X max – X min

The grand mean is the mean of the subgroup means:

j = number of subgroups

j

X

X

j

1 j j∑

=

=

Note: to establishstatistical process

control, 30 subgroupsof sample size n = 5 aretypically required (eight

subgroups are used

here for simplification)

©2009 ASQ 28

Steps for Calculating X-bar and R5. Calculate the mean of the ranges. R-bar divided by a constant d2 is an

unbiased estimate of the population standard deviation. The constant d2

is part of the constant A2.

6. Find the appropriate A2 value from the table of Control Chart Constants.

7. (See CSSBB Handbook Appendix 4 and 5; pages 463-464)

8. Calculate the UCL and LCL for the Mean

9. Find the appropriate D4 and D3 values from the table of Control Chart

Constants

10. Calculate the UCL and LCL for the Ranges

The mean range is the sum of the ranges divided by the

number of ranges:

j

R

R

j

1 j

j∑=

=

26.361.550.57725.47

RAXUCL 2Mean

=⋅+=

⋅+=

24.581.550.57725.47

RAXLCL 2Mean

=⋅−=

⋅−=

3.281.552.114

RDUCL 4Range

=⋅=

⋅=

0.01.550.0

RDLCL 3Range

=⋅=

⋅=

Note: if the number ofsubgroups is less than20, the table is incorrect

8/22/2019 08 CM0471 Module 8


Slide 26

RK5 what is count pointing to? It looks like it is pointing to a blank are b/t c chart and u chart. Is thatcorrect? Also, is there a way to move that bar to the right side of the slide, so it's not on the tan barrunning along the bottom of the sldie (I don't know how to rotate an arrow if it can be done).Robert Kraus, 12/21/2009

8/22/2019 08 CM0471 Module 8


©2009 ASQ 29

Steps for Calculating X-bar and R

11. Finished

©2009 ASQ 30

Graphing X-bar and R ControlCharts

1. For X-bar Chart, tick off equally spaced subgroups on the X-axis andtick off the UCL centerline (X-double-bar) and the LCL on the Y-axis

2. Use the Grand Mean to draw the Centerline

3. Draw lines for the UCL and LCL

4. Plot the Subgroup Means

5. Connect each point to form a scatter plot with connect lines

6. For R Chart, tick off equally spaced subgroups on the X-axis and tickoff the UCL, centerline (X-double-bar), and the LCL on the Y-axis

7. Use the Range Mean to draw the Centerline

8. Draw a line for the UCL and LCL

9. Plot the Subgroup Ranges

10. Connect each point to form a scatter plot with a connect line

11. Complete the X-bar and Range Chart (display them together)

©2009 ASQ 33

Mean and Standard Deviation(X-bar and S) ChartIn the average and standard deviation chart (X and S), thestatistic used to estimate the subgroup spread is the subgroupstandard deviation.

8/22/2019 08 CM0471 Module 8


©2009 ASQ 34

Steps for calculating X-bar and S(X-bar and Standard Deviation)1. Collect Data by Subgroup. The data shown are collected in

10 subgroups of sample size n = 5.

2. Estimate the population mean for each subgroup.

3. Estimate the population standard deviation for each subgroup.

4. Calculate the Grand Mean.

5. Calculate the mean of the subgroup standard deviations.

6. Find the appropriate A3 value from the table of ControlChart Constants.

7. Calculate the UCL and LCL for the mean.

8. Find the appropriate B4 and B3 values

from the table of Control Chart Constants.

9. Calculate the UCL and LCL forstandard deviation.

10. Finished.

See CSSBB HB,

pages 364 and365 for formulae

©2009 ASQ 35

Steps for Graphing X-bar and S

1. Follow the steps described for the X-bar and R chart.

2. Substitute standard deviation (S) for range (R) whereappropriate.

10987654321

220

210

200

190

180

Sample

S a m p l e M e a n

_ _ X=199.99

UCL=223.24

LCL=176.74

10987654321

40

30

20

10

Sample

S a m p l e S t D e v

_ S=23.85

UCL=40.93

LCL=6.77

Xbar-S Chart of Subgroup Data

©2009 ASQ 39

Individual and Moving Range(I and MR) Chart

8/22/2019 08 CM0471 Module 8


©2009 ASQ 40

Steps for Calculating I-MR(Individual and Moving Range)

1. Col lect data.2. Calculate the moving range between each pair of successive

values.

3. Calculate the mean for the data.

4. Find E2 in the table of Control Chart Constants.

5. Calculate the mean (MR-bar) of the moving range (MR).

6. Calculate the UCL and LCL for the individual observations.

7. Find D3 and D4 in the table of Control Chart Constants.

8. Calculate the UCL and LCL for the moving

range (MR).

9. Fin ished. See CSSBB HB, page 366for formulae

©2009 ASQ 41

Steps for Graphing I-MR Charts

1. For I Chart

• tick off equally spaced intervals for subgroups onthe X-axis.

• tick off equally spaced intervals that include the UCL,centerline (X-bar), and the LCL on the Y-axis.

2. Use X-bar to draw the centerline.

3. Draw lines for the UCL and LCL.

4. Plot the individual observation, X, for each subgroup.

5. Connect each point to form a scatter plot with connectlines.

©2009 ASQ 42

Steps for Graphing I-MR Charts

6. For the MR Chart

• tick off equally spaced intervals for subgroups on theX-axis

• tick off equally spaced intervals that include the UCL,the centerline (MR-bar), and the LCL on the Y-axis.

7. Use MR-bar to draw the centerline.


9. Plot the moving range, MR, for each subgroup.

10. Connect each point to form a scatter plot with connect lines.

11. Complete the I and MR Chart (display I Chart and the MRChart together).

8/22/2019 08 CM0471 Module 8


©2009 ASQ 46

p-Charts

©2009 ASQ 47

Steps for Calculating p-Charts

1. Record data by sample.

2. Calculate the sample totals.

3. Calculate fraction defective (proportion defective) p foreach sample.

4. Calculate the centerline, p-bar, the average fractiondefective (average proportion defective).

5. If one UCL and one LCL are desired (no moving limits),

calculate average sample size.

6. Calculate the UCL and LCL.

7. Finished.See CSSBB HB, page 368for formulae

©2009 ASQ 48

Steps for Graphing p-Charts

1. For the p Chart

• tick off equally spaced intervals for subgroups on theX-axis

• tick of equally spaced intervals that include the UCL,centerline (p-bar), and the LCL on the Y-axis

2. Use p-bar to draw the centerline.


4. Plot the fraction defective, p, for each subgroup.


8/22/2019 08 CM0471 Module 8


©2009 ASQ 52

np-Charts

©2009 ASQ 53

Steps for Calculating np-Charts

1. Collect the data.

(Example: There are 10 subgroups of 1000 pieces each. Total Defective = 376)1. Calculate the total defectives and total sample size.2. Calculate p-bar.

3. Calculate the centerline, np-bar.4. Calculate the UCL and LCL.

5. Fin ished.

0.037610,000/376Size)Sample(Total/)Defectives(Totalp ===

55.6 0.376)-1(0.03761,000337.6

)p - (1p[n3pnUCL

=⋅+=

+=

19.6 0.376)-1(0.03761,000337.6

)p - (1p[n3pnUCL

=⋅−=

+=

See CSSBB HB, page 370

for formulae

©2009 ASQ 54

Steps for Graphing np-Charts

1. For the np Chart

• Tick off equally spaced intervals for subgroups on theX-axis

• Tick off equally spaced intervals that include the UCL,centerline, and LCL on the Y-axis

2. Use np-bar to draw the centerline.

3. Draw lines for UCL and LCL.

4. Plot the number of defectives by subgroup.

5. Connect each point to form a scatter plot with connect lines.

8/22/2019 08 CM0471 Module 8


©2009 ASQ 57

c and u-Charts

With c- and u-Charts, it is very important to define

the defects and the unit.Examples:

Inspection Unit Type of Defects Counted

50 milesof pipeline

Weld defects

10 yards of cloth Blemishes, snags

50 circuit boardsSolder joint defects, damagedcomponents

100 forms Incorrect data entry, missing data

©2009 ASQ 61

c-Charts

©2009 ASQ 62

Steps to Calculate c-Charts

1. Collect data by subgroup

2. Calculate the total defects

3. Calculate the centerline, c-bar

4. Calculate the UCL and the LCL

5. Finish

See CSSBB HB, page 372for formulae

8/22/2019 08 CM0471 Module 8


©2009 ASQ 63

Steps for Graphing c-Charts

1. For the c Chart

• Tick off equally spaced subgroups on theX-axis

• Tick off equally spaced intervals that included

the UCL, the centerline (c-bar), and the LCLon the Y-axis

2. Use c-bar to draw the centerline.

3. Draw lines for the UCL and the LCL.

4. Plot the number of defects by subgroup.

5. Connect each point to form a scatter plot with

connect lines.

©2009 ASQ 66

u-Charts

©2009 ASQ 67

Steps for Calculating u-Charts

1. Collect data by subgroup: Shipping #, Rolls, and InspectedDefects. Calculate Defects per Roll.

2. Calculate the total rolls shipped.

3. Calculate the total defects.

4. Calculate the centerline, u-bar.

5. Calculate the UCL and LCL for an average sample size of n=1.

38.014

532 u ==

56.5 38.0338.0

1)(n

u3uUCL

=+=

=

+=

19.5 38.0338.0

1)(n

u3uUCL

=−=

=

−=

8/22/2019 08 CM0471 Module 8


©2009 ASQ 68

Steps for Calculating u-Charts

51.1 19.0338.0

2)(n

u3uUCL

=+=

=

+=

24.9 19.0338.0

2)(n

u3uUCL

=−=

=

−=

1. Calculate the UCL and LCL for an average sample size of n=2.

2. Finish.

©2009 ASQ 69

Steps for Graphing u-Charts

1. For the u Chart

• Tick off equally spaced intervals for the subgroups onthe X-axis

• Tick off equally spaced intervals that include the UCL,the centerline (u-bar), and the LCL on the Y-axis.

2. Use u-bar to draw the centerline

3. Draw the UCL and the LCL corresponding to the samplessize for the subgroup (n = 2 for subgroups 5 and 6; n = 1

for all other subgroups).

4. Plot the defects per unit by subgroup.


©2009 ASQ 72

Short Run SPC• Short-run or low-volume production is common in

manufacturing processes that produce built-to-order productsor quick turnaround production.

• The short-run control chart can also be used in other industriessuch as general services and healthcare when data arecollected infrequently.

• These processes often are so short that not enough data canbe collected to construct standard control charts.

• Statistical process control techniques have been developedto accommodate short-run production for both variables dataand attributes data.

• If possible, collect at

least 20 samples beforeyou construct the controlcharts for shortproduction runs.

See CSSBB HB, pages 376-382 for some examples.

8/22/2019 08 CM0471 Module 8


©2009 ASQ 79

Moving Average and Moving RangeCharts (MAMR)

•See CSSBB HB, pages 383-388for some examples.

•See page 383 for formulae

The moving average and moving range (MAMR)charts provide a graph of the moving average ofa process characteristic and the moving range.Moving averages are common in the financialindustry to analyze stock prices. The 200-day

moving average is the average closing price forthe past 200 trading days. With each new closingprice, a past price is excluded from the data.

©2009 ASQ 80

MAMR Charts

Key Points about MAMR Charts• Use with variable data

• Alternative to Shewhart X and R control charts• May be suitable when it is necessary to detect smaller process

shifts

• Appropriate to use when periodically collecting data• Relevant when it may be desirable to dampen the effects of

over control• Tends to smooth data, easier to spot trends

• Use with stable process mean

• Use time-ordered data• The selection of the moving average length affects the overall

sensitivity of the MAMR chart to detect process shifts.Generally, the longer the length, the less sensitive the chart isto detecting shifts – and, it follows, that a MAMR Chart with alength of two is less sensitive than and I and MR Chart.

©2009 ASQ 89

Progress Check

Check all statements that are true for MA charts:

• You can use MA charts when there are long timeintervals between consecutive results

• You can use MA charts when results are difficult ortime-consuming to obtain

• You can use MA charts when plotted points arecalculated from artificial subgroups created fromconsecutive observations

• The moving averages are independent

8/22/2019 08 CM0471 Module 8


©2009 ASQ 90

Interpret control charts and distinguish betweencommon and special causes using rules fordetermining statistical control. (Analyze)


VIII.A.5 Control Chart Analysis

©2009 ASQ 91

Control Chart Analysis

©2009 ASQ 92


Rule 1:Any point beyond the control limits is evidence thatthe process is not operating in a state of statistical

control.

8/22/2019 08 CM0471 Module 8


©2009 ASQ 93


Rule 2:

The second rule is known as a run. Seven or eightpoints in a row above or below the centerline isevidence of a special cause (and therefore theprocess is not stable).

©2009 ASQ 94


Rule 3:The third rule is called a trend. Six successive

points in an upward or downward direction isevidence that the process is not stable.

©2009 ASQ 95


8/22/2019 08 CM0471 Module 8


©2009 ASQ 96


©2009 ASQ 97


Example 1 – Process out of control

©2009 ASQ 98


Example 2 – Process in control

37332925211713951

8

7

6

5

4

3

2

1

0

Sample

S a m p l e C o u n t

_ C=2.775

UCL=7.772

LCL=0

C Chart of Blemish

8/22/2019 08 CM0471 Module 8


©2009 ASQ 99


Example 3 – Process cycling

37332925211713951

0.08

0.07

0.06

0.05

0.04

0.03

0.02

0.01

0.00

Sample

P r o p o r t i o n

_ P=0.02575

UCL=0.07327

LCL=0

P Chart of Ding

©2009 ASQ 100


Example 4 – Process drifting

37332925211713951

20

15

10

5

0

Sample

S a m p l e C o u n t

__ NP=12.05

UCL=21.82

LCL=2.28

NP Chart of Index

©2009 ASQ 101


Example 5 – Process shifting

37332925211713951

0.30

0.25

0.20

0.15

0.10

0.05

Sample

P r o p o r t i o n

_ P=0.1772

UCL=0.2918

LCL=0.0627

P Chart of M1

8/22/2019 08 CM0471 Module 8


©2009 ASQ 102

Progress Check

©2009 ASQ 103

Progress Check

©2009 ASQ 104

Lesson 2 – Other Control Tools

Define the elements of TPM and describe how it canbe used to control the improved process. (Understand)

VIII.B.1 Total productive maintenance (TPM)

8/22/2019 08 CM0471 Module 8


©2009 ASQ 105

Total Productive Maintenance (TPM)

Total Productive Maintenance

• TPM is a systematic approach for continuousimprovement of maintenance activities having

an impact on the control of a process.

• This strategy’s main goal is to maximize equipmentusefulness across its lifespan.

• TPM increases the Overall Equipment Effectiveness(O.E.E.), a combination of the uptime, cycle timeefficiency, and quality output of the equipment:

– O.E.E. % = Uptime% x Speed% x Quality%

• Note: Speed is also called Efficiency

©2009 ASQ 106

OEE Example: Which shift is better?

–O.E.E. % = Uptime% x Speed% x Quality%

See CSSBB HB, page 401for example.

Total Productive Maintenance (TPM)

©2009 ASQ 107

Progress Check

Calculate OEE, given the following data:

• Downtime = 15%

• Speed = 89%

• Quality = 93%

8/22/2019 08 CM0471 Module 8


©2009 ASQ 108

Define the elements of a visual factory and describehow they can help control the improved process.(Understand)


VIII.B.2 Visual Factory

©2009 ASQ 109

Visual Factory

• Setting up the workplace with signs, labels, color-coded markings, etc. to increase the awarenessof personnel working in different work areas andmultiple shifts to ensure consistency in a process.

• Visual aids help reduce variation in the processwhich can ultimately lead to defects.

©2009 ASQ 110

• Example: The branch network and mail operationsof a bank implemented a color-coded paymentsystem where each type of payment would beassigned a colored bag to ensure proper processing.

Visual Factory

8/22/2019 08 CM0471 Module 8


©2009 ASQ 111

Progress Check – Optional

Discuss the benefits of the following Visual Controls

in a workplace:

• Process metrics are typically displayed at the

machine or cell.

• Andon lights on a machine

• Work instructions posted in the production areas

• Proper placement of parts and tools

©2009 ASQ 112

Lesson 3 – Maintain Controls

Review and evaluate measurement system capabilityas process capability improves, and ensure that

measurement capability is sufficient for its intended

use. (Evaluate)

VIII.C.1 Measurement System Re-analysis

©2009 ASQ 113

Measurement System Re-analysis

• After implementing solutions, we must measure theprocess and determine if it has been improved to astatistically significant degree.

• Various tools such as measurement systemanalysis, process capability analysis, graphicaldata analysis, and statistical testing are used to

answer the question, “Did the improvements have

a statistically significant impact?”

• Statistically validated outcomes will help

demonstrate our process improvements.

• Measurement system capability is re-assessed toensure it remains adequate for the ever-improving

process capability.

8/22/2019 08 CM0471 Module 8


©2009 ASQ 114

Measurement System Re-analysis

Generally unacceptable; every effort should bemade to identify and correct the problem.Customers should be involved in determininghow the problem will be resolved.

Total measurement error of more than 30% oftotal tolerance

Possibly acceptable based on the importanceof the application, cost of the measuringequipment, cost of repairs, etc.

Total measurement error of 10% to 30% oftotal tolerance

Acceptable measuring equipmentTotal measurement error of less than 10% oftotal tolerance

Acceptability% of Measurement Error to Total Tolerance

Guidance for Acceptable ranges of Gauge R&R

©2009 ASQ 118

Develop a control plan for ensuring the ongoing

success of the improved process, including thetransfer of responsibility from the project team tothe process owner. (Apply)

Lesson 3 – Maintain ControlsVIII.C.2 Control Plan

©2009 ASQ 119

Control Plan

• “A control plan is a document describing the systemelements to be applied to control variation ofprocesses, products and services in order to

minimize deviation from their preferred values.”This living document summarizes the necessaryinformation used to explain, monitor and controla product or process and should be updated as

control methods are evaluated and improved.

– Glossary and Tables for Statistical Quality Control ,ASQ Statistics Division

•See CSSBB HB, page 406 for examples.

8/22/2019 08 CM0471 Module 8


©2009 ASQ 120

Control Plan Elements

•Control plans formats vary within industries,organizations, and even departments.•The format or template used for the control plan

is not the key factor; the elements included are.•The following are typical elements and examplesincluded in a control plan.

©2009 ASQ 121

Control Plan Elements - AIAG

“Header information” for the

part/processes being controlled

“Process information” for eachassociated process step

• P lan number

• Prototype/pre-launch/production

• Part/process number

• Suppl ier/plant

• Key contact person, phone

• Team

• Plant approvals, date

• Date developed, last revised

• Customer approvals, e.g.,

• Engineering, date

• Quali ty, date

• O ther , dat e

• Part process step number

• Process name/description

• Machine/device/tools for mfg.

• Characteristics – For eachproduct/process feature,

reference drawing/FMEA/ spec source

• Special “critical/safety/key” issues

• Specification tolerance

• Eval/measurement technique

• Sample size, frequency

• Control method

• Reaction plan

©2009 ASQ 122

Control Plan

Multi-disciplined team

Customer-to-Customer Circle

Product Planning

Product Development Manufacturing

Field Operations

Customer SatisfactionCustomer Satisfaction

Quality, Cost, and DeliveryQuality, Cost, and Delivery

• Gain an understanding of theprocess:

• Cause-Effect matrix

• QFD diagram

• FMEA

• Process maps

• Customer/technical requirements

• Others….

• Develop entries for Control Planform:

• Sources listed above, and…

• Fault tree analysis

• DOE results

• Results of BB studies

• Update Control Plans as needed:

• Process/spec changes

• BB improvements

8/22/2019 08 CM0471 Module 8


©2009 ASQ 123

Progress Check

A company delivers products to the customer for daily “Just inTime” requirements. Each day, the clerk takes several large

containers of products from inventory, loads them on the truck,hauls them to the customer, fills the customer’s smaller containersfrom the large ones, completes the paperwork, and then returnsunneeded quantities to the original inventory. Normally, customerdemand is stable and predictable, but the customer may call for adifferent product mix or large volumes for peak demands.

Assignment: As a group, work together and develop a control planfor the five-step process of :

1. loading the truck with the product2. unloading products into the customer’s JIT bins3. completing the paperwork for delivery/invoicing/inventory4. returning to stock products not left with the customer

5. revising the inventory balance

©2009 ASQ 124

Lesson 4 – Sustain Improvements

Document the lessons learned from all phases ofa project, and identify how improvements can be

replicated and applied to other processes in theorganization. (Apply)

VIII.D.1 Lessons learned

©2009 ASQ 125

Lessons Learned

• A Six Sigma project generates a wealth of information.

• Establishing a process to capture, document, and sharelessons learned infuses change in the organization.

• Document the knowledge and experience gained bycarrying out various projects in an organization.

• Ask and document the answers to the followingquestions:

– What went well?

– What could have been done differently?

– What could be improved?

– What did we do that we should not have?

8/22/2019 08 CM0471 Module 8


©2009 ASQ 126

Lessons Learned

– Did all our various stakeholders interact

efficiently and effectively?

– Where were the gaps?

– Where were the overlaps?

– What can be done differently next time to make

the situation easier for all parties involved?

• Once captured, project information can be searchedand compared against future project opportunities

• Decisions can then be made to charter a project,replicate a past success, or kick-start innovativethinking on a new or an existing project

©2009 ASQ 127

Progress Check

• As a team, discuss the benefits of conducting a“Lessons Learned” from all the phases of a SixSigma project

©2009 ASQ 128

Develop and implement training plans to ensurecontinued support of the improved process. (Apply)


VIII.D.2 Training Plan Deployment

8/22/2019 08 CM0471 Module 8


©2009 ASQ 129

Training Plan Deployment

Initial Training:

• This type is used for existing employees who havebeen responsible for executing the old process and

employees who are new to the process.

• It may be conducted either off-the-job, on-the-job,or both.

• Criteria is based on skill levels required, processcomplexity, and experience levels, among otherfactors.

• Initial training serves to calibrate employees and to

minimize variation in how the process is performed.

©2009 ASQ 130

Recurring Training

• This type is used to minimize deterioration in the

process performance over time.

• Specific facts, details, and nuances are frequentlyleft out as knowledge is passed on from worker

to worker.

• Restores the process execution to its original design.

• The frequency of recurring training should bedetermined based on process metrics and employeeperformance.

• Such training may be offered at specified intervalsor conducted as required to serve the needs of

underperforming employees.


©2009 ASQ 131

Considerations when developing training plans:

• Providing employees with the minimum skillsand information needed to perform the functions

required by the position.

• Providing employees with additional skills and

information that will assure a broader view of

what the position accomplishes for the enterprise.

• Providing employees with cross-training for

additional functions.


8/22/2019 08 CM0471 Module 8


©2009 ASQ 132

Other Considerations:

• Providing employees with opportunities for furthereducation outside the enterprise on topics not

directly related to the current organization needs.

• Providing incentives and requirements that motivateemployees to continue to seek education and

training opportunities.

• Providing employees with experiences thatdemonstrate the need the organization has for

their continuous improvement ideas.

• Providing employees with the opportunity to helpformulate a customized annual training plan.


©2009 ASQ 133

Progress Check

• As a team, discuss the method used by yourorganization for deploying training.

• What are some of the things that the organization

could consider when developing and deployingtraining?

©2009 ASQ 134

Progress Check

• Write down and discuss the “three most importantthings you learned at work” that is related to your job performance:

• How did you learn them? Next to each thing youlearned, note what you did to learn it. Do you know?Did it occur in a classroom? What was the reasonthat triggered the training?

8/22/2019 08 CM0471 Module 8


©2009 ASQ 135

Develop or modify documents, including standard

operating procedures (SOPs), work instructions, etc.,

to ensure that the improvements are sustained overtime. (Apply)


VIII.D.3 Documentation

©2009 ASQ 136

Documentation

Considerations for developing and updatingdocuments:

• Documented standard operating procedures (SOPs)

and work instructions help reduce process variation.

• The purpose of these documents is to make certainthat the activity is performed the same way over

time.

• This is especially important when multi-skilled,cross-trained personnel move into a variety of

positions.

©2009 ASQ 137

DocumentationConsiderations for developing and updating documents:• The development and updating of these documents must

involve the people who perform the work.

• Documents must be kept current.• Employees need to have access to the appropriate

documentation based on effectivity of the change.• Multiple formats for documentation exist (SOPs/WI/

templates, etc.).• Choose the right format for documentation.

– The right choice depends on how the documentation is tobe used, by whom, and at what skill level.

– Develop documents that utilize pictures, graphics, and arelight on words – or no words when there is more than onelanguage in the work force.

– Level of detail provided in any set of documentation shouldbe reflective of the skills and education levels of thepersonnel doing the actual work and the degree to whichvariation must be controlled.

8/22/2019 08 CM0471 Module 8


©2009 ASQ 138

Progress Check

• As a team, discuss the various types of documents

that are controlled by your organization.

• What is the method used to ensure that personnelare using the most effective version required for thework being performed?

©2009 ASQ 139

Identify and apply tools for ongoing evaluation ofthe improved process, including monitoring for new

constraints, additional opportunities for improvement,etc. (Apply)


VIII.D.4 Ongoing Evaluation

©2009 ASQ 140

Ongoing Evaluation

Tools such as control charts, process capabilitystudies, and process metrics:

• Control charts – These are used to monitor thestability of the process, determine when specialcause is present, and when to take appropriateaction.

• Process capability studies – These studies provideus with the opportunity to understand how the voiceof the process (i.e., control limits) compares with the

voice of the customer (i.e., specifications).

• Process metrics – This includes a wide variety of in-process and end-of-process metrics that measure

its overall efficiency and effectiveness.

8/22/2019 08 CM0471 Module 8


©2009 ASQ 141

Progress Check

As a group, discuss how your organization selects

and uses a tool for ongoing evaluation, once theprocess has been improved.

©2009 ASQ 142

Module Status

1. Statistical process control (SPC)

2. Other control tools3. Maintain controls4. Sustain improvements

©2009 ASQ 143

Module 8

Exercise Solutions

8/22/2019 08 CM0471 Module 8


©2009 ASQ 144

Answer

Which of the descriptions in the table below are

examples of Statistical Process Control? Check theappropriate box at the right.

Description/Name Example Non-Example

100% inspection (screening) of allproduction

X

Recording when assignable (special) causesare detected and taking corrective action

periodically (say, weekly)

X

Determining the natural variation of a

process

X

Assigning design engineers to maintain SPC X

Using samples to determine process stabil ity X

©2009 ASQ 145

Answer A candy-making process uses 40 pistons to deposit 40

chocolate pieces in a 5 x 8 array on a moving sheet of waxpaper. Below are two options that illustrate how a rational

subgrouping of sample size n = 5 may be selected:Option 1: the first five chocolates in each row formed by five

different pistons.Option 2: the upper left-hand chocolate formed in five

consecutive arrays by the same piston.

Check off the optimal box.Use a multi-vari chart to determine

where the variation exists: piece-to-piece(Option 1) or within piece (Option 2)

Answer: Option 2:

©2009 ASQ 146

Answer

Check all statements that are true for MA charts:

• You can use MA charts when there are long timeintervals between consecutive results. True

• You can use MA charts when results are difficult ortime-consuming to obtain. True

• You can use MA charts when plotted points arecalculated from artificial subgroups created fromconsecutive observations. True

• The moving averages are independent. False

8/22/2019 08 CM0471 Module 8


©2009 ASQ 147

Answers

Answers:1. D

2. C3. C

©2009 ASQ 148

Answers

Answers:

4. B5. B

©2009 ASQ 149

Answer

Calculate OEE, given the following data:

• Downtime = 15%

• Speed = 89%

• Quality = 93%

Answer:

O.E.E. % = Uptime% x Speed% x Quality%

If Downtime = 15%; then Uptime = 100-15 = 85%

OEE = .85* .89 * .93 = .7053 = 70.35%