I.3 Philosophy and History

I.3 Philosophy and I.3 Philosophy and HistoryHistory

Some Goals of ManufacturingSome Goals of Manufacturing Various ParadigmsVarious Paradigms Some Aspects of Taguchi’s Philosophy Some Aspects of Taguchi’s Philosophy

– On-Line and Off-Line Quality ControlOn-Line and Off-Line Quality Control– Design AspectsDesign Aspects

Some GoalsSome GoalsThe Quality TimelineThe Quality Timeline

1980

1920

1940

Manufacturing Process Design Product Design

Inspection

Process ControlQuality Control

Design Improvements

ParadigmsParadigmsThe Inspection ParadigmThe Inspection Paradigm

Inspect Output To RemoveInspect Output To RemoveUnsatisfactory ProductUnsatisfactory Product

Process Output Monitor

ParadigmsParadigmsShewhart’s ParadigmShewhart’s Paradigm

““Can’t Inspect Quality Into A Product”Can’t Inspect Quality Into A Product” Detect And Correct Quality Problems By Detect And Correct Quality Problems By Monitoring The Manufacturing Process Monitoring The Manufacturing Process

Two Types Of Variation In A Two Types Of Variation In A Manufacturing ProcessManufacturing Process

Variation ACV CCV= +

ParadigmsParadigmsShewhart’s ParadigmShewhart’s Paradigm - - TerminologyTerminology

Shewhart: Assignable Cause Shewhart: Assignable Cause Variation (ACV) and Chance Causes Variation (ACV) and Chance Causes Of Variation (CCV)Of Variation (CCV)

Deming: ACV=Local Faults, Special Deming: ACV=Local Faults, Special Causes, CCV=Common Cause Causes, CCV=Common Cause VariationVariation

Juran:Juran: ACV=Sporadic Problems ACV=Sporadic Problems CCV=Chronic CCV=Chronic ProblemsProblems

Variation ACV CCV= +

ParadigmsParadigmsShewhart’s ParadigmShewhart’s Paradigm

ObjectiveObjective

Monitor The Process To Detect When ACV Is Present.Monitor The Process To Detect When ACV Is Present.– How Do You Detect ACV? How Do You Detect ACV?

Control ChartsControl Charts

Process Output Monitor

Control

ParadigmsParadigmsShewhart’s ParadigmShewhart’s Paradigm

Two Ways To Improve QualityTwo Ways To Improve Quality Eliminate Assignable Causes To Eliminate Assignable Causes To Bring Process Into Statistical Bring Process Into Statistical ControlControl

Improve The System By Reducing Improve The System By Reducing Common Cause VariationCommon Cause Variation

ParadigmsParadigmsTweaking The Variation Paradigm:Tweaking The Variation Paradigm:The Variation/Complexity ParadigmThe Variation/Complexity Paradigm

Variation ACV CCV= +

Variation = Controllable Variation

+ Uncontrollable Variation

Complexity of the

Process= Unnecessary

Complexity + Inherent Complexity

ParadigmsParadigmsThe Variation/Complexity ParadigmThe Variation/Complexity Paradigm

Paraphrase Deming - “Reduce Paraphrase Deming - “Reduce Variation And Complexity And Variation And Complexity And You Improve Quality”You Improve Quality”

ParadigmsParadigmsDeming and JuranDeming and Juran

Originally Found That 80 to 85% Originally Found That 80 to 85% Of The Quality Problems Of The Quality Problems Required Management’s Input In Required Management’s Input In Order To Correct.Order To Correct.

Over The Years This Percentage Over The Years This Percentage Increased To Where They Found Increased To Where They Found That 94 to 96% Of Quality That 94 to 96% Of Quality Problems Were Systemic!Problems Were Systemic!

ParadigmsParadigmsDeming And Juran’s ExtensionDeming And Juran’s Extension

Need Management’s Need Management’s InvolvementInvolvement

Process Output Monitor

ControlManagement

ParadigmsParadigmsDeming and TaguchiDeming and Taguchi

Customer Satisfaction is the Customer Satisfaction is the Determining Factor in Product Determining Factor in Product DesignDesign– Taguchi - Quality is measured in terms Taguchi - Quality is measured in terms of loss to societyof loss to society

-Wring Out the Variation-Wring Out the Variation– Deming and Juran - Controlling the Deming and Juran - Controlling the Process (On-Line QC)Process (On-Line QC)

– Taguchi - Design Robust (Rugged) Taguchi - Design Robust (Rugged) Processes and Products (Off-Line QC)Processes and Products (Off-Line QC)

ParadigmsParadigmsTaguchi - Another Way To Improve QualityTaguchi - Another Way To Improve Quality

Build Rugged/Robust Products Build Rugged/Robust Products And Processes By Making Things And Processes By Making Things Insensitive To Unknown Or Insensitive To Unknown Or Uncontrollable Sources Of Uncontrollable Sources Of VariationVariation

Some Aspects of Taguchi’s Some Aspects of Taguchi’s PhilosophyPhilosophy

Prevention Rather Than DetectionPrevention Rather Than Detection Rather Than Detecting Defects, Let’s Design Quality Into The Rather Than Detecting Defects, Let’s Design Quality Into The Product Right From The Beginning.Product Right From The Beginning. “An Ounce of Prevention “An Ounce of Prevention...”...”

Process Output Monitor

ControlManagement

EngineeringCustomer

Some AspectsSome AspectsTaguchi’s ApproachTaguchi’s Approach

During the 50’s and 60’s During the 50’s and 60’s Developed A Philosophy and Developed A Philosophy and Methodology Concerning Product Methodology Concerning Product Design and ProductionDesign and Production

Some AspectsSome AspectsDesign Aspects of Taguchi’s PhilosophyDesign Aspects of Taguchi’s Philosophy

Robust Products And Processes Robust Products And Processes Through “Engineering Design” Through “Engineering Design” (Statistical Design Of (Statistical Design Of Experiments, DOE, Plays A Role Experiments, DOE, Plays A Role Here)Here)– Parameter DesignParameter Design

Choose Parameter (Factor) Levels ToChoose Parameter (Factor) Levels To– Optimize Process or Product PerformanceOptimize Process or Product Performance– Mitigate Uncontrollable FactorsMitigate Uncontrollable Factors

Some AspectsSome AspectsDesign Aspects of Taguchi’s Philosophy:Design Aspects of Taguchi’s Philosophy:

FactorsFactors

Control FactorsControl Factors– ““Easily” controlled sources of variationEasily” controlled sources of variation

NoiseNoise– UncontrollableUncontrollable

External - EnvironmentalExternal - Environmental

Goal - Design Product to Be Goal - Design Product to Be Insensitive to Noise FactorsInsensitive to Noise Factors