Six Sigma Innovation (DMAIC)

NNOVATION

CUSTOMER & COMPETITIVE INTELLIGENCE FOR

PRODUCT, PROCESS, SYSTEMS & ENTERPRISE

EXCELLENCE

I

DEPARTMENT OF

STATISTICSDR. RICK EDGEMAN, PROFESSOR & CHAIR – SIX SIGMA BLACK BELT

[email protected] OFFICE: +1-208-885-4410

IXS IGMA

S

IXS IGMAS

DEPARTMENT OF

STATISTICS

a highly structured strategy for acquiring, assessing, and applying customer, competitor, and enterprise intelligence for the purposes of product, system or enterprise innovation and

design.

Six Sigma COPIS Model

Customers Suppliers

Outputs InputsProcess

Steps

The Voice of the Customer (VOC) is aggressively sought and rigorously evaluated and used to determine

needed outputs and hence the optimal process configuration needed to yield those outputs and their

necessary inputs for which the best suppliers are identified and allied with.

From Concept to Market: the Voice of the Customer

How does Six Sigma Work?

TheVillain

level DPMO CP3

2 308,537 Not Applicable3 66,807 25%-40% of sales4 6,210 15%-25% of sales5 233 5%-15% of sales6 3.4 < 1% of sales

Each sigma shift provides a 10% net income improvement

Cost of Poorly Performing Processes

(CP3)

Why is Six Sigma Important?

Define

Control

Improve Analyze

Measure

Six Sigma Innovation

Define the problem and customer requirements.

Measure defect rates and document the process in its current incarnation.

Analyze process data and determine the capability of the process.

Improve the process and remove defect causes.

Control process performance and ensure that defects do not recur.

Six Sigma DMAICInnovation Approach

Six Sigma DMAICInnovation Approach

Slides 4 Through 32 of theProject Definition & Team Formation

Powerpoint File Represent the Six Sigma

DEFINE Stage

Slides 4 Through 32 of theProject Definition & Team Formation

Powerpoint File Represent the Six Sigma

DEFINE Stage

Define

Control

Improve Analyze

Measure







Measure: What Measurements are Important and

What Tools Should be Used?

1. Select Customer Critical to Quality (CTQ) Characteristics;

2. Define Performance Standards (Numbers & Units); 3. Establish the Data Collection Plan, 4. Validate the Measurement System,

5. and Collect the Necessary Data.

Measure:

1. Select Customer Critical to Quality (CTQ) Characteristics.

Among useful quality tools in the MEASURE phase are:

Quality Function Deployment (QFD) which relates CTQs to measurable internal sub-processes or product characteristics.

Process Maps create a shared view of the process, reveals redundant or Unnecessary steps, and compares the “actual” process to the ideal one.

Fishbone Diagrams provide a structure for revealing causes of the effect.

Pareto Analysis provides a useful quantitative means of separating the vital few causes of the effect from the trivial many, but require valid historical data.

Failure Modes and Effects Analysis (FMEA) identifies ways that a sub-process or product can fail and develops plans to prevent those failures. FMEA is especially useful with high-risk projects.

Kano Customer Need Model

Delighted

Disgusted

AbsentFullyImplemented

Stakeholder Satisfaction

TIME

Degree ofExecution

Measure: 1. Select Customer Critical to Quality (CTQ) Characteristics.

FAILURE MODES AND EFFECTS ANALYSIS (FMEA)

Failure Modes and Effects Analysis (FMEA) Process is a structured approach that has the goal of linking the FAILURE MODES to an EFFECT over time for the purpose of prevention. The structure of FMEA is as follows:

Preparation FMEA Process Improvement

a. Select the team b. Develop the process map and steps c. List key process outputs to satisfy internal and external customer requirements d. Define the relationships between outputs and process variables e. Rank inputs according to importance.

Measure: 1. Select Customer Critical to Quality (CTQ) Characteristics.



a. Identify the ways in which process inputs can vary (causes) and identify associated FAILUREMODES. These are ways that critical customer requirements might not be met.

b. Assign severity, occurrence and detection ratings to each cause and calculate the RISK PRIORITY NUMBERS (RPNs). c. Determine recommended actions to reduce RPNs. d. Estimate time frames for corrective actions. e. Take actions and put controls in place. f. Recalculate all RPNs.

FAILURE MODE: How a part or process can fail to meet specifications. CAUSE: A deficiency that results in a failure mode sources of variation EFFECT: Impact on customer if the failure mode is not prevented or corrected.

RATING DEGREE OF SEVERITY LIKELIHOOD OF OCCURRENCE

ABILITY TO DETECT

1 Customer will not notice the adverse effect or it is insignificant.

Likelihood of occurrence is remote.

Sure that the potential failure will be found or prevented before reaching the next customer.

2 Customer will probably experience slight annoyance.

Low failure rate with supporting documentation.

Almost certain that the potential failure will be found or prevented before reaching the next customer.

3 Customer will experience annoyance due to slight degradation of performance.

Low failure rate without supporting documentation.

Low likelihood that the potential failure will reach the next customer undetected.

4 Customer dissatisfaction due to reduced performance.

Occasional failures. Controls may not detect or prevent the potential failure from reaching the next customer.

5 Customer is made uncomfortable or their productivity is reduced by the continued degradation of the effect.

Relatively moderate failure rate with supporting documentation.

Moderate likelihood that the potential failure will reach the next customer.

6 Warranty repair or significant manufacturing or assembly complaint.

Moderate failure rate without supporting documentation.

Controls are unlikely to detect or prevent the potential failure from reaching the next customer.

7 High degree of customer dissatisfaction due to component failure without complete loss of function. Productivity impacted by high scrap or rework levels.

Relatively high failure rate with supporting documentation.

Poor likelihood that the potential failure will be detected or prevented before reaching the next customer.

8 Very high degree of dissatisfaction due to the loss of function without a negative impact on safety or governmental regulations.

High failure rate without supporting documentation.

Very poor likelihood that the potential failure will be detected or prevented before reaching the next customer.

9 Customer endangered due to the adverse effect on safe system performance with warning before failure or violation of governmental regulations.

Failure is almost certain based on warranty data or significant DV testing.

Current controls probably will not even detect the potential failure.

10 Customer endangered due to the adverse effect on safe system performance without warning before failure or violation of governmental regulations.

Assured of failure based on warranty data or significant DV testing

Absolute certainty that the current controls will not detect the potential failure.

FMEA Standardized Rating System1 < RPN = (Degree of Severity)*(Likelihood of Occurrence)*(Ability to Detect)

< 1000

Process or Product Name:

Prepared by: Page ____ of ____

Responsible: FMEA Date (Orig) ______________ (Rev) _____________

Process Step/Part Number Potential Failure Mode Potential Failure Effects

SEV Potential Causes

OCC Current Controls

DET

RPN

Actions Recommended Resp.

What are the process steps?

In what ways can the process step go wrong?

What is the impact of the Failure Mode on the customer?

How

sev

ere

is th

e ef

fect

on

the

cust

omer

? What are the causes of the Failure Mode?

How

ofte

n do

es th

e C

ause

or

Fai

lure

Mod

e oc

cur? What are the existing controls

and procedures that prevent the Cause or Failure Mode?

How

wel

l can

you

det

ect

the

Cau

se o

r F

ailu

re

Mod

e?

Cal

cula

ted What are the actions

for reducing the occurrence, decreasing severity or improving detection?

Who is responsible for the recommended action?

0

0

0

0

0

0

0

0

0

0

0

Failure Modes and Effects Analysis(FMEA)

Measure: 1. Select Customer Critical to Quality

(CTQ) Characteristics.



Develop and implement plans to reduce RPN’s.

Measure: 2. Define Performance Standards: Numbers & Units

At this stage customer needs are translated into clearly defined measurable traits.

OPERATIONAL DEFINITION: This is a precise description that removes any ambiguity about a process and provides a clear way to measure that process. An operational definition is a key step towards getting a value for the CTQ that is beingmeasured.

TARGET PERFORMANCE: Where a process or product characteristic is “aimed”If there were no variation in the product / process then this is the value that would always occur.

SPECIFICATION LIMIT: The amount of variation that the customer is willing to tolerate in a process or product. This is usually shown by the “upper” and “lower” boundary which, if exceeded, will cause the customer to reject the process or product.

DEFECT DEFINITION: Any process or product characteristic that deviates outside of specification limits.

Measure: 3. Establish Data Collection Plan, Validate the Measurement System, and

Collect Data.

A Good Data Collection Plan: a. Provides clearly documented strategy for collecting reliable data; b. Gives all team members a common reference; c. Helps to ensure that resources are used effectively to collect only critical data. The

cost of obtaining new data should be weighed vs. its benefit. There may be viablehistorical data available.

We refer to “actual process variation” and measure “actual output”: a. what is the measurement process used? b. describe that procedure c. what is the precision of the system? d. how was precision determined e. what does the gage supplier state about: f. Do we have results of either a:

* Accuracy * Precision * Resolution * Test-Retest Study? * Gage R&R Study?

Measure: 3. Establish Data Collection Plan, Validate the Measurement System, and Collect Data. Note that our measurement process may also have variation.

a. Gage Variability:

Precision: Accuracy: Both:

b. Operator Variability: Differences between operators related to measurement.c. Other Variability: Many possible sources. Repeatability: Assess effects within ONE unit of your measurement system, e.g.,

the variation in the measurements of ONE device. Reproducibility: Assesses the effects across the measurement process, e.g., the variation between different operators.Resolution: The incremental aspect of the measurement device.

Measure: 3. Establish Data Collection Plan, Validate the Measurement System, & Collect Data.GAGE R&R (Repeatability & Reproducibility) STUDY: a. Operators – at least 3 recommended;b. Part – the product or process being measured. It is recommended that at least 10

representative (reflects the range of parts possible) parts per study, with eachoperator measuring the same parts.

c. Trial – each time the item is measured. There should be at least 3 trials per part,per customer.

Source of Variation % ContributionTotal Gage Repeatability & Reproducibility

R1 + R2

Repeatability R1

Reproducibility R2

Part-to-Part 100% - (R1 + R2)

Total Variation 100%

Define

Control

Improve Analyze

Measure







Analyze: Where are we now? Where are we going?

What can prevent us from reaching our goals?

At this stage we determine the process sigma level and regardvariation as an enemy. We must determine process capability,

that is, the ability of the process to meet customer requirements.

We require several “z-scores” to make this evaluation.

ZBENCH Zst ZLT ZLSL ZUSL

Where: “BENCH” = benchmark, “st” = short term, “LT” = long term,

“LSL” = lower specification limit,and “USL” = upper specification limit.

Analyze: Where are we now? Where are we going? What can prevent us from reaching our goals?

ZST = best performance that can be expected from a process

ZLT = allows for drift through time (1 to 2 sigma drift is typical)

ZLSL= (X – LSL) / S then determine PLSL(d)

ZUSL= (USL – X) / S then determine PUSL(d)

P(d) = PLSL(d) + PUSL(d) then apply inverse use of the Z-table to findZBENCH (long-term)

P(d) * 1,000,000 = DPMO or PPM

0.0X

Zx.y P(d) ZBENCH

Analyze: Where are we now? Where are we going? What can prevent us from reaching our goals?

ZSHIFT = ZST – ZLT drift over time (DPMO tables assume 1.5)

ZST = (Specification Limit – Target) / ST * process sigma is determined here * indicates potential process performance if only common cause variation is present.

ZLT = (Specification Limit - ) / LT

* reveals long-term process capability * used to estimate DPMO or PPM

(“parts per million” same as DPMO) * includes special cause variation

^

^

Analyze:An Alternative Means of Approximating the Sigma Capability for

Your Process

StepAction Equations Your Calculations

1 What process do you want to consider? N/A Billing & Charging

2 How many units were put through the N/A 2,000process?

3 Of the units that went into the process, N/A 1,800how many were OK?

4 Compute process yield (step 3)/(step 2) 0.90005 Compute defect rate 1.0 – (step 4) 0.10006 Determine the number of potential N = number of 16

things that could create a defect critical-to-quality characteristics

7 Compute the defect rate per CTQ (step 5)/(step 6) 0.00625characteristic

8 Compute DPMO (step 7)*(1 million) 6,250

9 Convert DPMO to value conversion chart About 4.010 Draw conclusions JUST ABOUT INDUSTRY AVERAGE

Analyze: Where are we now? Where are we going?

What can prevent us from reaching our goals?

Yield Rates

A yield rate is a pass rate and can be characterized in various ways:

Classical Yield Rate = YC = (total defect free parts) / (total parts)

First Time Yield Rate = YFT = (parts defect free on the first pass)/(total parts)

Throughput Yield = YT = e–DPU where “DPU” = “defects per unit” and is calculated asDPU = (number of defects at any stage) / (total inspected). Note that due to rework some

items may be inspected multiple times with each inspection adding to the “total”.

Where are we now? Where are we going?What can prevent us from reaching our goals?

Yield RatesClassical Yield Rate = (total defect free parts) / (total parts) = ¾ = 75%

First Time Yield Rate = (parts defect free on the first pass) / (total parts) = ¼ = 25% Throughput Yield = e–DPU where “DPU” = “defects per unit” and is = e–18/8 = .1054

The “rework” that it takes to raise “throughput yield” to the “classical yield” level is called

Hidden Factory.

X

Symbol MeaningDefect

Meets Specifications

X Scrapped

Analyze

Analyze: Setting Performance Objectives

Critical to the Setting of Performance Objectives are the Concepts of‘Baseline’, ‘Process Entitlement’, ‘Benchmark’ and ‘Benchmarking’

BASELINE: This is the process performance level at the start of theSix Sigma Project.

PROCESS ENTITLEMENT: This is our best expectation for processperformance (e.g., the ‘sigma level’) with the current technology –

that is, without substantial reengineering or investment. This can be estimated from Zst.

BENCHMARK: This is the current ‘best in class’ performance level.

BENCHMARKING: The process of finding the benchmark performancelevel and then matching or exceeding that performance.

Analyze: Sources of Variation

This is the search for the Vital X’s –the factors that drive the customer CTQs.

Various statistical and quality methods are useful in this effort.Among these are:

HYPOTHESIS TESTING, which can

Reveal Significant Differences in Performance Between Processes Validate Process Improvements

Identify Factors that Impact the Process Mean and Variation.

FISHBONE or CAUSE-AND-EFFECT DIAGRAMS

Analyze: Sources of Variation:

The Hypothesis Testing Algorithm

1. Formulate the Null and Alternative Hypotheses, H0 and HA.

2. Specify the Sample Size and Significance Level of the Test, n and 3. Determine Which Type of Test Should be Employed.

4. State the Critical Value(s) & the Test Statistic & Specify the Decision Rule. 5. Collect and Validate Process Data.

6. Determine the Calculated Value of the Test Statistic (Data Based)7. As Appropriate, Construct and Interpret Confidence Intervals.

8. Determine and Pursue a Course of Action.

Key Vocabulary: Type I and II Errors, and

Define

Control

Improve Analyze

Measure







Improve: The goal of the improve phase is to test sourcesof variation to determine which of these actually

cause process variation in the customer CTQ.

7. Screen / Identify Causes of Variation. 8. Discover Variable Relationships.

9. Estimate Operating Tolerances & Pilot Solutions.

Improve: 7. Screen / Identify Causes of Variation.At this stage we determine which factors will be changed to improve the CTQs.

In step 6 (MEASURE) we selected the “vital few x’s” for each CTQ (little y).

In step 7 we select an appropriate improvement strategy based uponcharacterizing x’s as either

operating parameters or critical elements.

Operating Parameters are x’s that change in amount, rather than beingreplaced with another type / kind.

Operating Parameters can be set toseveral levels to see how they affect the process Y.

Critical Elements are x’s that are typically changed in type or kind,rather than in amount. These x’s are not necessarily measurable on a specific scale.

Improve:

7. Screen / Identify Causes of Variation.

Having identified the pertinent operating parameters and / or critical elements,we would then review whether Design of Experiments (DOE) would be

appropriate and, if so, develop the appropriate design, called a screening design.

The screening design is used to validate or eliminate factors (i.e. x’s),but is not ordinarily able to determine the optimal settings of the x’s.

Important considerations include the number of factors, number of levelsof each, the ‘range’ of settings for each factor, replication,

randomization & whether to use blocking variables.

Improve: 8. Discover Variable Relationships.

GOAL: to determine the precise changes needed

It is common to apply Optimizing DOE at this point, to determinethe ‘best’ settings of the x’s. It is common to use fractional factorial

designs or central composite designs to accomplish this goal.

It is common to include “baseline conditions” among thefactor settings. We desire to determine the “transfer function” (the regression equation). In combination these are intended

to yield a proposed solution to achieve project objectives.

Important considerations include the testing budget, availablepersonnel, and time allotted for the study.

Improve: 9. Estimate Operating Tolerances & Pilot Solutions.

PURPOSE:to estimate the range of values for each vital x that will satisfy customer requirements.

CONCEPT:

IF we can characterize the x-Y relationship AND we know the requiredspecifications of Y, THEN the tolerances can be set for each x factor.

Specifications flow down from customer requirements and we adjust tolerancesaccounting for variation, unless variation is small enough to be ignored.

STATISTICAL TOLERANCING

Improve: 9. Estimate Operating Tolerances & Pilot Solutions.

YUSL

YLSL

YTarget

{

{

1

1

2

2

}

3}

3

xUxL

Y’USL

Y’LSL3

3This graph indicates an indirect relationship between x and Y

Original YSpecification

s

Adjust for Variationin Y, THEN

Adjust further forvariation in X.

Sta

tistical

Tole

ran

cing

Define

Control

Improve Analyze

Measure







Control: The Goal of the Improve Phase is to Test Sources of Variation to

Determine which of These Actually Cause Process Variation.

10. Validate the measurement system of the control variables. 11. Determine process capability.

12. Implement process control system & bring the processto a close.

Control: 10. Validate the measurement system of the control variables.

GOAL: Make sure the implemented solution remains effective and in control

HERE WE ESTABLISH THE Process Control System

Even though our solution may be an excellent one, the nature of most systems is toward entropy or degradation, thus we will: (a) Create an implementation plan with controls for each x (b) Prepare documentation and provide (for) training, and (c) Collect data to re-evaluate process capability.

RECALL our Measurement System Analysis, as used in Step 3 of MEASURE:a. what is the measurement process used? b. describe that procedure c. what is the precision of the system? d. how was precision determinede. what does the gage supplier state about: f. Do we have results of either a:

* Accuracy * Precision * Resolution * Test-Retest Study? or a * Gage R&R Study?

In MEASURE the MSA was applied to y. In CONTROL it is applied to x.

Control: 11. Determine process capability.

The GOAL at this stage is to statistically confirm that theimplemented changes have produced improved performance.

Process capability is reassessed.

A Hypothesis Test may be appropriate to evaluate the difference in performance prior

to and after the implemented changes

Control: 12. Implement process control system and bring the project to a close

Three primary approaches may be used at this stage:

Risk Management: This is similar to FMEA but now focus is trained on x, rather than y.

Risk Management Score = RMS = (Impact)*(Probability) RM identifies and quantifies risks, establishes a risk abatement plan,

and monitors the progress of the plan.

Mistake Proofing:This is a technique for eliminating errors by making it impossible to make them inthe process. To quote: “It is good to do it right the first time. It is even better

to make it impossible to do it wrong.”

Statistical Process Control (Charts): This is a feedback system with sequential data and ongoing process data collection.

Black Belt Perspective1 Select CTQ Characteristic2 Define Performance Standards3 Validate the Measurement

System

4 Establish Product Capability5 Define Performance Objectives6 Identify Variation Sources

7 Screen Potential Causes8 Discover Variable Relationships9 Establish Operating Tolerances

10 Validate Measurement System11 Determine Process Capability12 Implement Process Controls

Measure

Analyze

Improve

Control

Business Level View R Recognize the true states of your business.D Define what plans must be in place to realize improvement of each state.

M Measure the business systems that support the plans.A Analyze the gaps in system performance benchmarks.

I Improve system elements to achieve performance goals.C Control system-level characteristics that are critical to value.

S Standardize the systems that prove to be best-in-class.I Integrate best-in-class systems into the strategic planning framework.

Operations Perspective

R Recognize operational issues that link to key business systems.

D Define Six Sigma projects to resolve operational issues.

M Measure performance of the Six Sigma projects.

A Analyze project performance in relation to operational goals.

I Improve Six Sigma project management system.

C Control inputs to project management system.

S Standardize best-in-class management system practices.

I Integrate standardized Six Sigma practices into policies and procedures.

A Process ViewR Recognize functional problems that link to operational issues.

D Define the processes that contribute to the functional problems.

M Measure the capability of each process that offers operational

leverage.A Analyze the data to assess prevalent patterns & trends.

I Improve the key product / service characteristics created by the key processes.

C Control the process variables that exert undue influence.

S Standardize the methods & processes that produce best-in-class performance.

I Integrate standard methods & processes into the design cycle.

NNOVATIONI

DEPARTMENT OF

STATISTICSDR. RICK EDGEMAN, PROFESSOR & CHAIR – SIX SIGMA BLACK BELT

[email protected] OFFICE: +1-208-885-4410

IXS IGMA

S

End of Session

Documents

Six Sigma Innovation (DMAIC)