six sigma - rolul campionilor de proiect

7/27/2019 Six Sigma - Rolul campionilor de proiect

1/135

SIX SIGMA

Role of Project Champions

Material prepared by Robert H. Lochner, Ph.D.Copyrighted 2003, 2004


2/135

What is Six Sigma?


3/135

Six Sigma is:

A measure of quality of a process

An approach to process and systemdesign and management which leads to

superior quality and productivity


4/135

Six Sigma as a Measure of Quality

Sigma Percent Defects per Million

Level Defective Opportunities (DPMO)

6 0.000% 3.45 0.02% 320

4 0.4% 6,210

3 6.6% 66,8002 30.8% 308,000


5/135

Examples of Sigma Processes

6s: Domestic airline fatality rate.

6s:Anesthesia deaths during surgery.

4s: Preventable hospital deaths. 3s:Airline lost baggage

2s: Mammography screening

1s: IRS telephone tax advice

Source: GE Medical Systems, in article by W.H. Ettinger in Sept. 2001 Trustee.


6/135

Examples of Sigma Processes

There are 365x24x60 = 525,600 minutes per year.

6s: No water at faucet 1.8 minutes per year.

5s: Telephone not working 2 hours and 48 minutesper year.

4s: No electricity 54 hours a year.


7/135

Common Weaknesses of

Management Systems

Production schedule comes first, cost

second, quality third Inadequate or no quality planning

Quality viewed as a Production issue

Quality by inspection


8/135

Common Weaknesses of

Management Systems - continued

Customers and Production not involved innew product design

Lack of understanding of process variation

Inadequate training of employees

Decisions based on opinions

No understanding of concept of internalcustomer


9/135

Strategy for Success

Decisions should be:

Driven by what customers want, not what

the boss wants.

Based on data, not opinions.

Made with input from individuals closest tothe situation.


10/135

Evolution of Modern Quality

Inspection

Statistical Quality Control

Quality Assurance

Total Quality Management

Six Sigma


11/135

Inspection

Quality is assured by inspecting finished

products. The goal is uniform quality.Individuals are held responsible for thequality of finished products.


12/135

Statistical Quality Control

Statistical tools developed in the 1920sare used to produce uniform products with

reduced inspection. Unacceptableproducts are caused by out of controlprocesses. The Quality ControlDepartments job is to solve quality-related

problems. Controlling variation is a key issue


13/135

Quality Assurance

It is understood that superior products canbe produced only by cooperation of all

departments involved in design,production and delivery of products.Processes, not people, are the source ofmost quality problems. Quality is built into

products. Reducing variation is a priority.


14/135

Total Quality Management

Quality is a management issue, not amanufacturing issue.

Quality is a competitive issue.

Quality is driven by customer expectations

Quality Improvement actions must bebased on data.


15/135

W. Edwards Deming

Leading quality guru of 20th Century.

Credited with bringing modern quality to

Japan.

A physicist turned statistician, Demingstresses need to measure and reduce

process variation. Taught that quality was a management

issue.


16/135

Deming Chain Reaction

Improve Quality

Costs Decrease

Increased market shareStay in business and grow

Fulfill the missions of our sponsors


17/135

Dr. Demings Profound Knowledge

Knowledge of variation, types of variation,statistical theory.

Knowledge about tampering.

Knowledge about interaction of forces.

Knowledge of operational definitions.

Knowledge of psychology.


18/135

Definition of Six Sigma

Six Sigma is a comprehensive and flexible systemfor achieving, sustaining and maximizing

business success. Six Sigma is uniquely drivenby close understanding of customer needs,disciplined use of facts, data, and statisticalanalysis, and diligent attention to managing,

improving, and reinventing business processes.

from pg. xi ofThe Six Sigma Wayby Pande, Neuman and Cavanagh


19/135

In a Six Sigma organization:

Quality is a strategy for increasing marketshare and profitability.

Variation is reduced to levels previouslythought unattainable.

Modern statistical methods are used to

understand what drives processes andachieve breakthrough improvements.


20/135

Six Sigma:

Customer driven

Process focused

Prevention oriented

Decisions are based on data

Extensive use of statistical methods

Supported by infrastructure


21/135

Benefits of Six Sigma:

Significant cost savings

Increased customer satisfaction

Increased market share

Increases productivity

Reduces cycle times

Promotes cross-functional cooperation


22/135

Success Story - Motorola

Between 1987 and 1997, sales increasedfive-fold, profits increased 20% per year.

Total savings of $14 billion.


23/135

Success Story General Electric

Started Six Sigma in 1995, pay-off of $750million by end of 1998.

A team in billing cut invoice errors by 98% A team in GE Capitol streamlined contract

review process, saving $1 million per year

A GE Medical Systems team redesignedmedical scanning technology to reduce time forfull body scan from 3 minutes to 30 seconds


24/135

VARIATION


25/135

If I had to reduce my message for

management to just a few words, Id

say it all had to do with reducingvariation.

-- W. Edwards Deming


26/135

Variation

Every process has variation.

The amount of variation which can be

tolerated in a process depends on theimpact the variation has on the processand on customer satisfaction.

Excessive variation leads to errors, delays,scrap, rework and lost business.


27/135

Traditional Quality Loss Function

0

Loss tobusiness

Specificationlimits


28/135

Quadratic Quality Loss Function

0

Loss tobusiness

Specificationlimits

Target


29/135

Cost of Quality


30/135

Fixed Costs of Quality

Time employees routinely spend oninspection and verification of their work.

Storage of extra materials as safetyfactors

Extra production to allow for defects

Routine complexities, inefficiencies


31/135

Variable Costs of Quality

Overtime caused by quality problems

Time spent soothing unhappy customers

Damage to products in handling, spoilage

Expediting orders due to purchasingproblems

Lost business due to poor quality


32/135

Cost of Quality Categories

Appraisal

Internal Failure

External Failure

Prevention


33/135

Appraisal Costs

Traditional costs associated with qualitycontrol such as:

Inspecting incoming materials, finishedproducts, orders, invoices.

Monitor customer service staff.

Monitor suppliers.


34/135

Internal Failure Costs

Costs associated with unacceptable products orwork discovered prior to delivery to the customer

such as: Damaged supplies or products.

Rework and overtime due to quality problems.

Sorting out unacceptable products.

Extra work due to inaccurate forecasting orordering of materials.


35/135

External Failure Costs

Costs resulting from unacceptable productsor services being delivered to a customer

such as: Returns or adjustments due to errors or

defects.

Time spent handling customer complaintsand making adjustments.


36/135

Prevention Costs

Time and resources spent to preventproblems or improve operations such as:

Training in quality concepts and methods. Six Sigma project teams.

Quality planning.

Process capability analyses.


37/135

Quality Costs Exercise

A company has monthly sales of $1,000,000.

Profits are 6% of sales. Then profits = ____

Quality costs are 8% of sales. Then quality costs = ____

Failure costs are 75% of quality costs. Then failurecosts = ____

Suppose failure costs are reduced by 10%. Then profitswould increase by ____

How much more sales would be needed to realize thesame increase in profits as could be obtained bydecreasing failure costs by 10%? ____


38/135

Organizing for

Six Sigma


39/135

Six Sigma Roles

Management

Sponsor/Champion

Process Owner

Black Belt, Green Belt

Team Leader, Team Members

Implementation Leader


40/135

Management Responsibilities

Strong, visible, sustained leadership.

Incorporate Six Sigma into strategic plan.

Study Six Sigma concepts and methods. Select projects, black belts, green belts,

champions and team leaders.

Provide support infrastructure.

Conduct project reviews. Approve and support team recommendations.


41/135

Sponsor/Champion Role

Work with management in developing projectcharter and selecting team leader and members.

Meet regularly with Black Belt and team leaderto review project progress.

Assure adequate resources for team.

Support the team and the project withmanagement.


42/135

A Project Champion must:

Show sincere interest in and enthusiasm for theproject.

Tell others how important the project is.

Support but not direct the project. Anticipate problems.

Deal with roadblocks quickly and directly.

Learn about the Six Sigma approach to processimprovement and the tools teams use.


43/135

Champions should evaluate drivingforces and barriers to the Project

Driving Forces Barriers/Obstacles


44/135

Process Owner

Keep informed on progress of team.

Provide input and support to team.

Secure cooperation of employees working in theprocess.

Assist project team in obtaining information andwhen implementing improvements.

Process owner may also be the projectchampion and implementation leader.


45/135

Black Belt

Provide training to team on DMAIC and teamprocesses.

Provide training in use of process and data

analysis tools. Perform or lead data collection and statistical

analysis activities. Provide assistance to individuals on data-based

decision making. Coach management in Six Sigma concepts. Half-time to full-time responsibility.


46/135

Team Leader

Is familiar with process to be improved.

Has good interpersonal skills.

Conducts team meetings. May be a Black or Green Belt.

Works with Black/Green Belt and team membersoutside of meetings to keep project moving.

Provides progress reports to Champion, processowner and management.


47/135

Implementation Leader

Develops timeline and plan for implementingimprovements.

Assures appropriate documentation and training. Actively monitors the new process to prevent

slippage.

Usually the process owner is the implementationleader.


48/135

Project Team Participants

Leader is usually the process owner or, for yourprojects, a Black Belt trainee.

Team members must include people working in

the process. If possible, internal and/or external customers

and suppliers should be on team. Outside experts can be part of team or called in

as needed. Individuals can have more than one role.


49/135

DMAIC: The Six Sigma Approach to

Process Improvement

DefineaMeasureaAnalyzeaImproveaControl


50/135

What is DMAIC?


DMAIC is a five-step structured approach toprocess improvement and processdesign/redesign. It evolved from and is an

improvement upon the Deming Cycle.


51/135

DMAIC Cycle Step 1: Define

Identify the problem.

Determine customer requirements.

Define the scope of the project.

Set goals for the project.


52/135

DMAIC Cycle Step 2: Measure

Identify key input, process and outputmeasures

Measure current state of process;measure key inputs and outputs.

Validate problem/issue with data.

Compare current process measures tocustomer expectations and businessneeds.


53/135

DMAIC Cycle Step 3: Analyze

Discover root causes.

Identify causal relationships.

Identify key control factors.

Assess adequacy of current or proposedprocesses.


54/135

DMAIC Cycle Step 4: Improve

Develop strategy to improve process orrecommend new process.

Test proposed changes/new process. If proposed actions have desired effect,

implement them system-wide.

If proposed actions do not have desiredeffect, go to steps 1 or 2, as appropriate.


55/135

DMAIC Cycle Step 5: Control

Document new standard operatingprocedures.

Train employees in new process. Determine appropriate measures of

process performance.

Monitor ongoing process performance. Correct problems as needed.


56/135

Step 1: Define Opportunities



57/135

Caution:

One of the most common and costlyerrors that occur in Six Sigma

implementation is selection ofinappropriate projects.


58/135

External sources of project ideas

Voice of the Customer: What feedback haveyou gotten from customers regarding satisfactionwith current products/services and anticipated

future needs. Voice of the Market: Is your market share or

number of orders dropping? Are currentproducts becoming obsolete?

Comparison with Competitors: How are youperceived relative to your competition in quality,service, delivery, cost?


59/135

Internal sources of project ideas

Voice of the Process: How are key processesperforming? Are scrap rates, down time orovertime increasing or varying greatly? Are

schedules missed? Are there frequent costoverruns? What processes require constantmonitoring or adjustment?

Voice of the Employee: What do those closest toyour processes have to say about what is not

working well? Are there disagreements betweendepartments?


60/135

Criteria for project selection

Important to your customers

Part of your strategic plan

Opportunity for significant cost savings Chronic, difficult problem

Solution to problem unknown

Narrowly defined scope Doable in 4 months with available resources


61/135

Issues regarding project selection

Select core rather than support processes

Base selection on data, not opinions

Has management already decided what to do? Will the project enhance cross-functional

cooperation?

Is there a sense of urgency?

Is variation a major problem?


62/135

Project team charter should include:

Mission statement:Goal of the project

Measurable criteria of success Reason for project selection

Scope/boundaries of project

Constraints on time, cost and resources List of participants


63/135

Examples of a Mission Statements:

Reduce production down-time in final assemblyby 10%.

Reduce occurrence of missed medications in

nursing units by 30%. Ship 95% of all orders within 7 days of receiving

them.

Increase chemical process yield by 15%.


64/135

Examples of Reasons for Project

Selection: Lose of market share due to late shipments.

Rising customer complaints about finish on

motor housings. Unable to match price of competition due to high

production costs.

Extended hospital stays due to missed

medications.


65/135

Examples of Scope of Project:

Final assembly operations.

Responding to customer inquiries regarding

orders from receipt of call to answering customerconcerns.

From physician prescribing medication to patientreceiving medication.

From grain arriving on a truck until grain is instorage unit.


66/135

Examples of Resource Constraints:

Overall costs must not be increased.

Project must be completed within 150 days.

Staff from Information Services will assist asneeded.

Financial services will provide all financial datarequested by the team, subject to approval of

the team champion.


67/135

Customer

Requirements


68/135

Customer-Focused Quality

The customer decides what constitutesquality.

Customer expectations can be defined interms of:Cost

Delivery

Quality

Service


69/135

Customer Expectataions

Customers classify features and servicesinto three basic categories:

Basic requirementsIf theyre not presentthey are called dissatisfiers.

Satisfiers - features that are expected in

varying degrees Delighters - unexpected pleasing features


70/135

Example of customer expectations

A person goes to a restaurant for a chickensandwich.

Basic requirement the meat is cooked. Satisfier the meat is properly seasoned.

Delighter the bun is freshly baked.


71/135

Another example

A person feels sick and goes to see a doctor

Basic requirement doctor says he/she knows

what is the problem and recommends treatment. Satisfier You recover from the illness quickly.

Delighter The doctor gives you free medicineor calls the next day to see how you are doing.


72/135

Grid for identifying customerexpectations

Cost Delivery Quality Service

Basic

Satisfier

Delighter


73/135

Exercises Customer expectations

What are your expectations when you buy ashirt?

What range of room temperatures is comfortablefor you?

For your customers, or a specific subset ofcustomers, fill in as much as you can of the grid

on the previous slide. How would you get theinformation needed to complete the grid?


74/135

Customer Data

Qualitativeinformation like I liked the waythe shirt fit or there was too much salt on

the hamburger has limited usefulness. Quantitativeinformation like I had to wait

on the phone for 10 minutes or the

temperature in my hotel room was 20degrees last night is much more useful.


75/135

Quantifying Customer Expectations

Customers non-quantitative expectations,called the Voice of the Customer(VOC)need to be translated into measurablecharacteristics of the product or service,called Critical Customer Requirements(CCR). Critical to Quality measures (CTQ)

are specifications on process output whichneed to be met in order to achieve CCRs.


76/135

In other words . . .

If the process outputs meet their CTQs,

then the product or service meets its

CCRs, and the customer is happy because

his/her expectations (VOC) were met.


77/135

Example

VOC: I dont want to wait for my luggage.

CCR: Luggage to be available within 15

minutes of airplane arriving at the gate. CTQ: All luggage must be off the airplane

within 9 minutes of planes arriving at the

gate.


78/135

Sources of Customer SatisfactionInformation

Complaints

Salesmen

Warranty data Sales data

Surveys


79/135

Document the

Process


80/135

Basic SIPOC Diagram

SupplieraInputsaProcessaOutputsaCustomer

The Supplierprovides Inputs: materials, information orresources.

The Process is a series of steps that uses Inputs to produceOutputs.

The Outputs are then delivered to the Customer.

Reference: The Six Sigma Way by Pande, Neuman and Cavanagh, 2000, pages 168-173.


81/135

SIPOC Example: Shipping an Order

InputCheck

customercredit

Issuepacking

order

Pick &packorder

Label&

ship

Output

CompletedOrder form

BoxedProduct& Invoice


82/135

Top-down Process Flow Chart

What is it?

A SIPOC diagram with process sub-steps listed belowthe basic steps.

Why do we use it?

Major steps in the process are easy to identify.

It shows the relative complexity of different steps.

Sub-steps can be easily added later.


83/135

Example of Top-down Flow Chart:Shipping an Order

InputCheck

customer

credit

Issuepacking

order

Pick &pack

order

Label&

ship

Output

CompletedOrder form

BoxedProduct& Invoice

Checkcreditstatus*Problemson earlier

orders?

*Verifyin stock

*Verifyaddress

*Print

order

*Find items*Select box*Put itemsin box*Seal box

*Addresslabel

*ShippingInstructions

*Handlingwarnings


84/135

Traditional Process Flow Chart

Shows greater detail than SIPOC.

May have branches and decision points.

Suitable for use with simple processes orparts of larger processes.

Useful for identifying process complexity.


85/135

Example of Process Flow Chart

CheckCustomer

creditObtainorder

IssuePacking

orderCheckinventory

Adequate?

Pick &Packorder

Label&

ship

PrepareInvoice

Holdorder

NoYes


86/135

Process Flow Chart Symbols

Below are some symbols commonly used inprocess flow charts:

Process start and end

Action step

Decision point

Direction of movement


87/135

Block Diagram(Also called a functional deployment process map.)

What is it?A block diagram is a process flow chart in

which activities of different departments areput in separate rows or columns.

Why do we use it?

A block diagram highlights interfaces betweendepartments.

E l f Bl k Di


88/135

Example of Block Diagram:Preparing Product Information Sheet

Engineering

Marketing

ProductInformation

Production

Printing

Shipping

Obtaininformation

Createblueline

Review

Review

Review

Revise

Review

Review

Review

Clarify

Print

Inventory Ship


89/135

Process Map

What is it?A process map is a flow chart in which

measurable inputs, outputs and processcontrol factors are shown for each step.

Why do we use it?Six Sigma projects are data driven. A process

map shows us what measurements, and ofwhat type, are available for the project.


90/135

Xs and xs Ys

The output of any process (Y) isdetermined by the inputs into the process

(Xs) and the factors that control theprocess (xs):

Input: X Process Output: Y

X1, x2,, xk


91/135

Process Parameters

X: measurable process input

x: measurable factor affecting process

outcomesY: measurable final outcome of a process

y: measurable outcome of a process step


92/135

Pharmac

y

Nursing

Unit

Transmit

order to

Pharmacy

Ordergenerated

Prepare

medication

Transcribe

MAR andVerify with MD

order

AdministerMedication

CompleteDocumentation

Correctly

transcribed

MAR

Patient receives

correct medication

on time

Physician

order

Processcompleted

Info documented on MAR

Process Map if Medication Stored in Pharmacy

C Legibility of order

C Legibility of fax

N Allergies, BSA,

P Chart flagged

P Shift to shift

communication

N Patient specifies

times

P D/C orders

P Stop ordersP Order date/time

Verification of

completeness,

allergy,

interactions,

availability, BSA

C Fax machines working

C Unit clerk available

Legible order

N Volume of orders

P Patient info available

C Pharmacist available

C Dose scheduling

Drug added to patient profile

Patient info (allergies) into

Pharmacy data base

MAR label generated to

nursing unit

Product label generated

N Product available

N Mixing required

C Packaging

Apply label

Pharmacist

verify and send

C Printing equip

Labeled final

product

Correct medication to correct location,

labeled for correct person, on time

C Arrival time of MAR

N Volume of meds for

a patient

C RN available

C Unit clerk available

P No standard procedure

C/N Agency staff or floats

C Med clipboard available

N Deviation from routine

admin. times

C Adequate training

C Med available

C Nurse available

C Patient info sheet avail

N/C IV access

C Nurse familiar with med

N Patient available

N Patient refuses med

N Patient specifies times

N Route needs to be changed

N/C Form of drug changed

P Additive stock avail in unit

C MAR available

C Nurse available

C Duplicate charting

P Order of forms

C Variation in knowledge

of documentation process

C Pharmacist available

C Medication & ID correct

P Delivery method (tube, hand

deliver, server exchange

Unlabeled

final product


93/135

VARIATION


94/135

If I had to reduce my message for

management to just a few words, Id

say it all had to do with reducingvariation.

-- W. Edwards Deming


95/135

Variation

Every step of every process has variation.

The amount of variation which can be

tolerated in a process depends on theimpact the variation has on the processand on customer satisfaction.

Excessive variation leads to errors, delays,scrap, rework and lost business.


96/135

Example: Sony USA vs. Sony Japan

LSL USL

Sony JapanSony USA

Distribution of color density in television sets


97/135

Discussion

Why does variation increase costs?

What examples are there from your

company?


98/135

Types of Variation

When seeking to improve the performanceof a process, the first objective is usually

to reduce variation. There are two basictypes of variation:

Common cause variation,

Special cause variation.


99/135

Common Cause Variation

Common cause refers to the persistent variability whichis a part of any process. The amount of common causevariation present in a process is largely a function ofprocess design and management.

Common Cause Variation: Consists of many individual causes, each making a

small contribution to the variability. Varies in level from place to place and process to

process. Is present in every step of every process. Can usually be reduced only by management action

by changing the process.

E l f C C V i ti


100/135

Examples of Common Cause Variation

Natural variation in raw materials.

Minor vibrations in machinery.

Temperature fluctuations due to heater cycling.

"Normal" variation in times to complete productionorders.

"Normal" variations in plating thickness.

Errors in medical insurance claims form due to poorly

worded instructions. Typographical errors in a report due to poor lighting in

the office.


101/135

Special Cause Variation

Special cause refers to temporary or local disturbancesin a process which have a noticeable effect on one ormore process characteristics. Special causes areexceptions to, rather than part of, the system.

Special cause variation:

Is the result of one or a few individual causes.

Is much larger than common cause variation.

May appear and then leave the process sporadically.

Can often be eliminated through action by employeesworking in the process.


102/135

Examples of Special Cause Variation

Voltage surge which damages equipment

Excess materials due to a major order being cancelled

Product dropped on the floor

Broken gauge Shipment late due to a blizzard

Order lost in the mail

Errors in order due to inexperienced clerk

Machine settings entered incorrectly due to intoxicatedoperator

Relationship Between Common Cause


103/135

Relationship Between Common Causeand Special Cause Variation

Special cause variation is usually the result ofcommon cause variation which is present in a

larger system. Attempting to reduce common cause variability

through adjustments by employees is called"tampering."

Tampering usually leads to increased processvariation.

Dealing with Variation


104/135

Dealing with Variation

Special cause variation should be removed from the process.If that is not possible the process should be evaluated andpossibly redesigned to minimize the effects of this specialcause.

Common cause variation should be measured and evaluatedwith respect to specifications and customer expectations.

If common cause variation does not result in unacceptableproducts or services, the process should be maintained atthat level of variation. Employees working in the processshould look for opportunities to reduce the variation,however.

If common cause variability is so large that it results inunacceptable products or services, the process should beredesigned to eliminate the problem. This may involvemaking the process less sensitive to certain uncontrollablefactors.

St bl P


105/135

Stable Processes

Processes which have only common causevariation are said to be stable.

For a stable process measurements of process,product and service performance will vary overtime, but the distribution of measurements will

be constant over time.


106/135

Measuring

Performance

Step 2 of DMAIC Process

Why do we measure?


107/135

Why do we measure?

To ASSESSCompare with standards, requirements, benchmarks

Establish benchmarks

Identify opportunities for improvement

To IMPROVE Discover actions which will improve performance

Evaluate effects of improvement actions

To CONTROLMaintain process stability

Adjust to changing situations


108/135

Why measure outputs (y, Y)?

They are of immediate interest to customers.

They provide summary measures ofperformance.

They aid in identifying processes in need ofimprovement.

They aid in setting priorities for improvement.

They enable organizations to measure theeffects of implementing process changes.


109/135

Why measure process factors (X, x)?

Virtually all results can be changed only bychanging the processes that produce them.

By monitoring critical process factors it ispossible to assure desired results.

Process measures can sometimes be used assurrogates for output measures.


110/135

What should we measure?

Characteristics which are important tocustomers.

Characteristics which are important to thebusiness (costs, cycle times, scrap rates).

Characteristics which affect processes.


111/135

Operational Definition

An operationaldefinition is "one peoplecan do business with." W. Edwards Deming

An operational definition is a clear,unambiguous definition which enables usto make measurements." William A.

Golomski


112/135

Operational definitions of specifications and of

customer requirements include some way of judgingwhether those specifications and requirements havebeen met. For example, what does "eighty pounds ofsugar" mean? Is 79.6 pounds close enough? If a batchis to be mixed for ten minutes at 90oC, will there be a

problem if the batch is mixed for 10 minutes and 15seconds? If it is mixed for only 9 minutes? And whereand how should the temperature of the batch bemeasured? What are the standards for neatness inpackaging, testing package seals? What is the definition

for acting "in a courteous manner"? What constitutes agood job as a first-line supervisor?

Fi P i i l f M


113/135

Five Principles of Measurement

1. Know what you are measuring.

2. Let customers tell you which outputcharacteristics to measure.

3. Ask continuously how your competitorsare doing.

4. Track your processes.

5. Tell your people everything you know.


114/135

Using measurement

to find root causes


115/135

We begin now the process of asking why?

We dig for root causes. First we obtain Y and yvalues, and determine whether they consistentlysatisfy our CTQ requirements. If they dont wewant to find the CTPs -- the input (X) and

control (x) factors which affect these outcomemeasures. We will now look at three processtools which help us discover potential rootcauses of unacceptable variation.

f f C


116/135

Tools for Identifying Potential CTPs

Pareto Chart

Cause-and-Effect Diagram

Is/Is Not Analysis

P t Ch t


117/135

Pareto Chart

What is it?

A bar chart which graphically displays

frequencies of various types of problems. Why use a Pareto chart?

It helps prioritize quality problems.

It can be used to present non-numeric data.

C C


118/135

Customer Complaints in April

0

5

10

15

20

25

30

35

40

45

50

Containe

rdam

aged

Toospicy

Prod

uctnotu

nifo

rm

Flavo

rtoo

stro

ng

Offta

ste

Toodry

Oth

er-notp

rodu

ct

Off

colo

r

Notsp

icyeno

ugh

Toobl

and

Complaint Count

Container damage 43

Too spicy 29

Product not uniform 18

Flavor too strong 17

Off taste 12

Too dry 11

Other not product 8

Off color 8

Not spicy enough 3

Too bland 2

C d Eff t Di


119/135

Cause-and-Effect Diagram

What is it?

A graphical tool which aids in discovering themost likely root causes of a problem orcondition.

Cause-and-effect diagrams are also calledIshikawa diagrams (after their developer) and

fish bone diagrams (after their appearance).

C d Eff t Di ti d


120/135

Cause-and-Effect Diagram, continued

Why do we use it? Provides structure to brainstorming. The creation process itself is educational.

Discussions of diagrams helps team membersunderstand the entire process being reviewed.

Facilitates the active search for root causes. Aids in seeing relationships among causes. Demonstrates the group's level of understanding.

The more complex the diagram, the moresophisticated the participants are in their knowledgeof the process.

Cause-and-effect diagram for errors by data


121/135

entry operators:

Taiichi Ohno former Toyota Motor Vice President used the


122/135

Taiichi Ohno, former Toyota Motor Vice President, used thefollowing example of looking for the root cause of a machinebreakdown (which is what cause-and-effect diagrams are all

about):Question 1: Why did the machine stop?Answer 1: Because the fuse blew due to an overload.Question 2: Why was there an overload?Answer 2: Because the bearing lubrication was inadequate.

Question 3: Why was the lubrication not adequate?Answer 3: Because the lubrication pump was not working

properly.Question 4: Why wasn't the lubricating pump working?Answer 4: Because the pump axle was worn out.

Question 5: Why was the axle worn out?Answer 5: Because sludge got in.Question 6: Why...

I /I N t A l i


123/135

Is/Is Not Analysis

One structured form of stratification is Is/Is-NotAnalysis, developed originally by Charles H. Kepner andBenjamin B. Tregoe. This can be useful where extremestratification effects are noted -- that is where there are

significantly different scrap rates on different days orlines, or different patterns in field failures among differentcustomer groups, or where customer complaints showgeographic patterns. The Is/Is Not Matrix on the nextslide can be used as part of the analysis.

Reference: P.R. Scholtes, The Team Handbook; C.H. Kepner and B.B. Tregoe, The Rational Manager.

IS: Where, When, IS NOT: Where, ANALYSIS:What and for Who etc. does the How might we


124/135

does the situation situation NOT explain thisoccur? occur? pattern?

WHERE has the

problem or situationoccurred, or wherehas it been observed?WHEN did the problemoccur? Is its occurrencerelated to the occurrence of other events?WHAT type of situationoccurred? What wasthe extent or durationof the problem?WHO was present orinvolved with the pro-

cess when the situationoccurred?


125/135

Types of

Measurement Data

Categories of Measures


126/135

Categories of Measures

Discrete measures

Also called attributes measures

Continuous measures

Also called variables measures

Discrete Measures


127/135

Discrete Measures

Discrete measures refer to qualitative datathat can be categorized and counted.

Examples include the number of errors ininvoices, and the percent of ordersrequiring expediting, the number of times amachine was down for repair, the number

of under-weight packages.

Continuous Measures


128/135

Continuous Measures

Continuous measures refer to quantitativedata such as temperatures, weights, tear

strengths, completion or delay times, costsand distances. They are generally muchmore valuable for improving processesthan are attributes measures.

Continuous versus Discrete Measures


129/135

Continuous measures are generally much morevaluable for improving processes than arediscrete measures. However, it is far better tohave a meaningful discrete measure than have

a continuous measure which is not stronglyrelated to quality or customer satisfaction. Forexample "lost time" data that does not includethe reasons for the lost time may not be as

valuable as discrete data on occurrences ofspecific packaging problems.

Measure xs


130/135

Measure x s

Processes cannot be improved by measuringjust the characteristics (Ys) of the productsproduced by the process. What happens to the

product is a result of what is going on in theprocess. Management and quality improvementteams should focus on critical characteristics oftheprocess (xs), since if the process can be

held in proper control, then the finished productwill be acceptable.


131/135

Tools for Graphical

Data Analysis

Dotplot


132/135

p

Recommended for displaying data sets of up to 20measurements. Each measurement is represented by adot.

Example: A persons heart rate was measured on tendifferent occasions. Observed rates were 77, 76, 80,79, 85, 76, 86, 82, 74 and 82 beats per minute.

o o

o o o o o o o o70 72 74 76 78 80 82 84 86 88 90

Histogram


133/135

196.5194.5192.5190.5188.5186.5184.5182.5180.5178.5

9

8

7

6

5

4

3

2

1

0

Sodium

Freque

ncy

Stem-and-Leaf Diagram


134/135

Stem and Leaf DiagramThe stem-and-leaf diagram is a tool developed to replace

the histogram. It contains more information than thehistogram and is easier to make. The stem-and-leafdiagram is recommended for use with data sets having20 or more measurements and at least 15 distinctvalues.

0 00000011111223345558891 00122235592 001334445563 28994 09

5 04

HI 8.51 (#13), 10.88 (#28)

Boxplots


135/135

Boxplots

The boxplot is a tool used to compare graphically three ormore sets of data.

42.0

42.5

43.0

43.5

44.0

44.5

Measure

six sigma - rolul campionilor de proiect

Documents