Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 8

My goal in this course is to give you the training to perform at the Six Sigma Green Belt level.

We will
cover measurement systems, graphs and charts, descriptive statistics, hypothesis testing, design of
experiments, SBC charts, control plans, and much, much more

ASQ Green Belt Certified

Six Sigma projects are focused on addressing specific problems, performance deficiencies, or
improvement opportunities

Define Phase:

1. Types/Identification of projects

1.1. Implementation projects


1.2. Lean projects, also known as kaizen events, or workouts
1.3. DMAIC projects
1.4. DMADV projects

2. Collecting VOC (Voice of the Customer)

3. Project Management basics. 

3.1. Project champion launches the projects


3.2. Communicate the project charter consistently
3.3. Develop Project plan: can use gantt charts to track project progress
3.4. Focus on Key metric Y [y=(x)]
3.5. Document Project Deliverables
3.6. Monitor project progress (project leader should provide weekly updates to the
champion)
3.7. Plan tollgate end of phase reviews (DMAIC phases)
3.8. Celebrate Project success.

Measure Phase

1. Types of process maps

1.1. The high-level process map, 


1.2. The detailed process map
1.3. The swimlane map

2. Sampling in data collection

2.1. Sampling strategies to avoid


2.1.1.Judgement sampling
2.1.2.Convenience sampling
2.2. Good Sampling strategies
2.2.1.Simple random sampling
2.2.2.Stratified random sampling
2.2.3.Systematic sampling
2.2.4.Subgroup sampling.

3. Descriptive statistics

3.1. Measures of central tendency (mean, median, mode)


3.2. Measures of spread/ dispersion (range, variance, standard deviation)
4. Graphical displays of variation

4.1. Dotplots (continuous data)


4.2. Histograms (min 45 data points required) (continuous data)
4.3. Boxplots (continuous data)
4.4. Pareto charts (discrete/ categorical data)

5. Measurement system analysis (MSA)

5.1. Attribute agreement analysis


5.1.1. Percent agreement between appraisers > reproducibility
5.1.2. Percent agreement within appraisers > repeatability
5.1.3. Percent agreement with the expert
5.2. Gage repeatability and reproducibility (for continuous data)

6. Process performance measures

6.1. DPU (defects per unit)


6.2. DPO (defects per opportunity)
6.3. DPMO (defects per million opportunities)
6.4. RTY (rolled throughput yield)
6.5. COQ (cost of quality)
6.5.1. Prevention (planning, training & design reviews)
6.5.2. Appraisal (inspections)
6.5.3. Internal failure (within org- defects, scrap & rework)
6.5.4. External failure (warranty claims & recalls)
6.6. COPQ (cost of poor quality)
6.6.1. Appraisal (inspections)
6.6.2. Internal failure (within org- defects, scrap & rework)
6.6.3. External failure (warranty claims & recalls)

Analyze phase

1. Hypothesis testing phase

1.1 Type of Hypothesis


1.1.1 Null Hypothesis H0
1.1.2 Alternate Hypothesis Ha
1.2 Level of significance (alpha) (5%)
1.3 Level of confidence (beta) (95%)
1.4 P Value (if P <= alpha, reject H0, if P> alpha, fail to reject H0)
alpha + beta = 1
2. Hypothesis testing protocol

3. Tests for means (assuming data follows normal distribution and variances are equal)

3.1 Anova: use to comparing means of more than 2 groups


3.2 One sample T test: use to compare mean of 1 group to a target
3.3 Two sample T test: use to compare means of 2 groups

4. Tests for variances

4.1 One sample variance test


4.1.1 Chi-square method (if data is normally distributed
4.1.2 Bonett (if not or don’t know)
4.2 Test for equal variances (2 or more)
4.2.1 Bartlett’s test (if data is normal)
4.2.2 Levene’s test (if not)

5. Tests for proportions

5.1.1 One-proportion test


5.1.2 Tow-proportion test

6. Tests for independence (Chi-square test)

7. Correlation and linear regression


Improve Phase

1. Introduction to design of experiments (DOE) – use to derive best combination of Xs

1.1. Independent variables (Xs)


1.2. Dependent variables(Y)
1.3. 2^K Factorial design (k=no of Xs)

2. Introduction to design for six sigma (DFSS)- take care of customer needs.

2.1. DFSS Characteristics


2.1.1. Quality is “designed in.”
2.1.2. Customer expectations/needs are prioritized from the start.
2.1.3. Requirements flow down. (link btwn customer req & functional product and process req)
2.1.4. Quality is predictable.

2.2. DFSS Phase (use for design or redesign product/service/process and address customer
needs systematically)

2.2.1. Define
2.2.1.1. Establish business case for design projects
2.2.1.2. Approve project charter
2.2.2. Measure
2.2.2.1. Analyze voice of the customer (VOC)
2.2.2.2. Identify & prioritize critical to quality requirements (CTQs)
2.2.3. Analyze
2.2.3.1. Examine CTQs to determine functionality needed
2.2.4. Design
2.2.4.1. Create product features
2.2.4.2. Finalize specifications
2.2.4.3. Develop process requirements
2.2.4.4. Optimize parameters
2.2.5. Verify
2.2.5.1. Reach a final design consensus
2.2.5.2. Test for producibility
2.2.5.3. Design controls and implement

3. Use of selection matrices

3.1. Criteria selection matrix (scaling 1 to 10)


3.1.1. Establish evaluation criteria
3.1.1.1. CTQs
3.1.1.2. Impact on Y
3.1.1.3. Cost
3.1.1.4. Ease of implementation
3.1.2. Apply relative weights to criteria
3.1.2.1. Allocate (i.e.,10 = cost, time, impact on Y) points to criteria
3.1.3. Place solutions across top of matrix

3.1.4. Score each alternative against criteria with scale of 1 to 10

Tip: Anchoring the score

3.2. Pugh concept selection method (compared to the same baseline)


3.2.1. Better (+)
3.2.2. Same (S)
3.2.3. Worse (-)
4. How to use failure modes and effects analysis (FMEA)

4.1. Risk priority number RPN = Occurrence x Severity x Detection


Control phase

1. Statistical process control charts: used by operators/owners to control process, by indicating when
& when not to intervene

1.1. it focusses on variations


1.1.1. Random variation (common causes)
1.1.1.1. Poor design
1.1.1.2. Bad methods
1.1.1.3. Lack of training or supervision

1.1.2. Nonrandom variation (special causes)


1.1.2.1. Jammed equipment
1.1.2.2. Broken tools
1.1.2.3. Contaminations
1.1.2.4. Untrained temp workers

1.2. SPC Charts plot


1.2.1. Product dimensions
1.2.2. Processing times
1.2.3. Number of occurrences
1.2.4. Number of defects
1.2.5. Proportion of defects

1.3. Upper & lower control limits = Average value +/- 3 standard deviations

2. SPC charts for variables

2.1. Types of SPC charts

2.1.1. Variables using continuous data (dimension & time)


2.1.1.1. Individual chart (I chart): Individual value such as processing time for invoices)
2.1.1.2. Individual and moving range chart (I-MR chart): where moving range = current
value minus previous value

2.1.2. Attributes using discrete data (counts of defects & proportions) (Point 3.1)
2.1.3. Out of control conditions for variable SPC charts
2.1.3.1. A point outside of the control limits
2.1.3.2. 9 points in a row on the same side of the centerline
2.1.3.3. 6 points in a row, all increasing or all decreasing
2.1.3.4. 14 points in a row, alternating up and down
2.1.3.5. 2 out of 3 points, more than 2 standard deviations from the centerline on the
same side
2.1.3.6. 4 out of 5 points, more than 1 standard deviation from the centerline on the
same side
2.1.3.7. 15 successive points within 1 standard deviation of centerline either side
2.1.3.8. 8 successive points more than 1 standard deviation from centerline either side

3. SPC charts for attributes (where discrete data is used)

3.1. Types of Attributes SPC charts


3.1.1. P chart (plots proportion of defective units)
3.1.2. NP chart (plots number of defective units)
3.1.3. C chart (plots number of defects)
3.1.4. U chart (plots number of defects per unit)

3.2. Out of control conditions for Attributes SPC charts


3.2.1. Any point outside of control limits
3.2.2. 9 points in a row on same side of centerline
3.2.3. 6 points in a row, all increasing or decreasing
3.2.4. 14 points in a row, alternating up and down

4. How to develop control plans (Tips)

4.1. Control subject should include


4.1.1. Outcome Y
4.1.2. Proven Xs
4.1.3. CTQ requirements

4.2. Measure frequently to avoid adverse outcomes

4.3. Ensure authorized personnel are local enough to take action

You might also like