1. Acceptance Sampling
2. Statistical Process Control
3. Design of Experiments

1. 1928: by Dodge and Romig at Bell Labs
2. Large lot of N parts with D amount of defects; take small sample n and count the number of defects d in sample n; reject if d>c otherwise accept lot

1. 1925, W. A. Shewhart at Bell Labs
2. Critical statistics of quality characteristics are periodically measured and plotted on a control chart with an upper and lower limit, aka Shewhart control chart
3. It will signal excessive process change when the statistics value fall outside the pre-determined control limits

• Trained in engineering and physics
• Long Career at Bell Labs
• Developed first control charts in 1924

• 1922/1923 by R. A. Fisher for use in agriculture science
• Popularized for industrial use by George Box
• Taguchi added robust design on top of DOE
• DOE is to discover the key input variables or factors influencing the quality characteristics of interest in a process or product, through systematic experimentation.

Sir Ronald Aylmer Fisher

R.A. Fisher

• Pioneered the Taguchi method used to reduce costs and improve quality
• Quality Loss Function
• Robust Design

• Continual never ending improvement
• Met Walter Shewhart at Bell Labs

1. Improve Quality
2. Cost decrease
3. Productivity Increases
4. Capture the Market
5. Provide Jobs and more jobs

Deming's 14 Points

Part 1

1. Create constancy of purpose towards improvement
2. Adopt a new philosophy, in a time of change
3. Cease reliance on mass inspection
4. stop rewarding business on basis of price alone
5. improve constantly and forever
6. Institute training
7. Improve leadership, recognize supervision is help people and equipment do a better job

Deming's 14 Points

Part 2

1. Drive out Fear
2. Break down barriers between departments
3. Eliminate slogans and targets for the workforce
4. Eliminate work standards
5. Remove barriers that rob workers of the right to pride in the quality of their work
6. Institute a vigorous program of education and self-improvement
7. Put everyone to work to accomplish the transformation

• AKA Deming's Wheel
• Plan, Do, Check, Act

• Romanian, worker at Western Electric and influenced by Shewhart
• Emphasize a more strategic and planning-oriented approach to quality than Deming

1. Planning
2. Control
3. Improvement

American quality control expert and businessman

Invented Total Quality Management (TQM)

The early 1980s

Emphasis on widespread training, quality, awareness, and continuous improvement

Forms basis of all quality systems today but was unsuccessful at first because there was not enough emphasis on quality engineering statistics tools, poor follow-through, no project-by-project implementation strategy, and training by HR

Disciplined and analytical approach to process and product improvement

Specialized roles for people (Champions, master black belts, black belts, green belts)

Top-down driven

• Define problems in process
• Measure Performance
• Analyze causes of problems
• Improve process - remove variations and non-value-added activities
• Control process so problems do not recur

Committed Leadership, Use of top talent

Supporting Infrastructure: Formal project selection and review process, dedicated resources, financial system integration

Project-by-Project implementation strategy

Order release system used and large machine setup time exists (Often a push system)

Batches move from one stage to next

Often have large WIP and inventory levels

High setup and change-over costs

Difficult to balance workflow

Overburdened Employees

Large growth potential in non-value added activities which add to production costs

Father of the Toyota Production System

This became Lean Manufacturing in the U.S.

Quick Changeover/Setup Reduction

Pull/Kanaban System

Workplace Organization 5S System

Value Stream Analysis

Total Productive Mantainence

Cellular Manufacturing

Tradition View:

Revenue = Sales Price - Cost of Production

Lean View of Profit:

Sales Price = Revenue + Cost of Production

Focus on Elimination of Waste:

Long cycle times, long queues, inadequate throughput, rework, and non-value added work activities

Makes use of:

5 why's, root cause analysis, and FEMA

1. Overproduction
2. Waiting
3. Transport
4. Overprocessing
5. Inventory
6. Motion
7. Defects
8. Skills - Underutilized personnel resources and creativity

1. Sort
2. Stabilize
3. Shine
4. Standardize
5. Sustain

WIP=Th*CT

WIP=Works in Progress

Th=Throughput

CT=cycle time

Rate at which we need to produce

T/D; T=Time to produce; D=Demand

Cannot be measured can only be calculated

moving the mean

Reduces average cycle time and excess inventory

improves the average production rate

Improve speed, Sustain Performance, and Achieve breakthrough

Yields:

Higher quality, lower costs, culture change, and dashboard results

Design for Six Sigma: requirements allocation, Capability assessment, robust design, predictable product quality

Lean: Lead time, flow mapping, waste elimination, cycle time, work-in-progress reduction, and operations/design

Six Sigma/DMAIC: Variation reduction, predictability,feasibilityy, efficiency, capability, accuracy

1. Define
2. Measure
3. Analyze
4. Design
5. Verify

ISO 9000:2000 Fundamentals and Vocabulary

ISO 9001:2000 Requirements

ISO 9004:2000 Guidelines for Performance Improvement

Created by U.S. Congress in 1987

Recognizes U.S. corporation for excellence

5 categories: manufacturing, service, small business, healthcare, and education

Prevention Costs

Appraisal Costs

Internal Failure Costs

External Failure Costs

Legal Costs

sample size of n is selected without replacement from a population N items

k items can be classified as successes and N-k is failures

not a binomial because binomial requires the probability of success be constant for every trial

Consists of repeated trials

each trial can result in only two outcomes

probability of success is constant for every trial

trials are independent

experiment results in outcomes that can be classified as successes or failures

average number of successes that occurs in a specific region is known

probability that a success will occur is proportional to the size of the area

Special case of negative binomial distribution

Deals with number of trials required for a single success

Negative Binomial Distribution where number of successes (r) = 1

AKA as Pascal Distribution

Consists of x number of repeated trials with only two possible outcomes

Probability of success, p, is same for every trial

Trials are independent

Experiments continue until r number of successes are observed, where r is specified in advance

1. Histogram
2. Check Sheet
3. Pareto Chart
4. Cause and Effect Diagram
5. Defect Concentration Diagram
6. Scatter Diagram
7. Control Chart

Producer's risk, α error

P(product is reject | but product is good)

Consumer's Risk, β error

P(product is not reject | but product is bad)

Power = 1-β

P(Reject H₀ | H₁ true)

Cp<1

Cp=1

Cp>1

Not Capable

Marginally Capable

Capable

Cpk<1

Cpk=1

Cpk>1

Not Capable

Marginally Capable

Capable

Pp<1

Pp=1

Pp>1

Not meeting Specs

Marginally meeting specs

Meeting specs

Cp=Pp

Process is stable

Ppk<1

Ppk=1

Ppk>1

Not meeting specs

Marginally meeting specs

Meeting specs

Cp=Ppk

Process is stable and centered