Statistical Process Control

1
Statistical Process Control

• Establishing SCADA /SPC/SQC
• Objectives and a Roadmap

2
Agenda

• Introduction of Participants
• Introduction to Six Sigma Methodology
• SPC Terminology and Tools of the Trade
• Process Mapping
• Cost and Effect, Failure Modes Effect Analysis,
  Attribute and Variable Data
• Determining Process Capability
• What you need to know to apply SPC
• Building a Process Map, CE Matrix, defining
  Control Charts
• Action Plan Moving Forward

3
William P. Southard

• President of DST CONTROLS
• Mr. Southard is an electrical engineer and a
  registered professional engineer for control
  systems in the state of California.
• He founded DST and has applied his expertise for
  over 25 years in projects involving computers,
  drives, programmable controllers, and supervisory
  control systems.

4
Frank Napoleon, Jr.

• COO of DST CONTROLS
• Mr. Napoleon is an electrical engineer with 15
  years of experience in the automation industry.
• He is former North American Sales Manager for GE
  Fanuc Automation
• He is a certified Six Sigma "Green Belt,"
  indicating he has received 80 hours of structured
  training in methods to improve quality.

5
Agenda

• Introduction of Participants
• Introduction to Six Sigma Methodology
• SPC Terminology and Tools of the Trade
• Process Mapping
• Cost and Effect, Failure Modes Effect Analysis,
  Attribute and Variable Data
• Determining Process Capability
• What you need to know to apply SPC
• Building a Process Map, CE Matrix, defining
  Control Charts
• Action Plan Moving Forward

6
If you have never

• Been forced to produce beyond the limits of
  process capability
• Argued with a customer about whose gauge is
  right
• Sorted parts to meet a specification
• Had too much scrap, rework, inspection
• Cursed a late payment, a wrong bill, a discrepant
  delivery
• Produced too much or too little
• Faced a cost reduction that couldnt be done

then youve probably already implemented Six
Sigma
7
Six Sigma Quality Premise

• All processes have variability
• All variability has causes
• Typically only a few causes are significant
• To the degree they can be understood, they can be
  controlled
• Designs must be robust to the effects of the
  remaining process variation

8
What about ISO9000?

• This Standard Describes Four Facets of Quality
• Quality due to definition of need for the
  product the quality due to defining the product
  to meet marketplace needs and opportunities.
• Quality due to product design the quality due to
  designing into the product the ability to meet
  the market need, and provide value to the
  customers.
• Quality due to conformance to the design the
  quality due to maintaining day-to-day consistency
  with the product design.
• Quality due to product support the quality due
  to furnishing support through the product life
  cycle as needed.

9
Statistical Process Control

• Applies to part three, Conformance to the Design
• It is about achieving day-to-day consistency with
  the product design
• Concerned with monitoring process output to give
  confidence that the process is
• Producing consistently satisfactory product
• Giving timely warning of the need for corrective
  action

10
Yf(x)

• Every process looks like this
• Strategyf(market conditions, et al
• Designf(requirements, et al
• Productf(materiallabor, et al

Process
In
Out
11
Reality
Pressure
Interruptions
function
Tool Wear
Operator
Day of the Week
Speed
Phase of the Moon
Yf(lots of Xs)
Variability
12
Distribution of Results
LSL
USL
s
- Product at spec
13
The Sigma Scale

• Z PPM Yield
• 2 308,537 69
• 3 66,807 93.3
• 4 6,210 99.3
• 5 233 99.98
• 6 3.4 99.9997

14
Agenda

• Introduction of Participants
• Introduction to Six Sigma Methodology
• SPC Terminology and Tools of the Trade
• Process Mapping
• Cost and Effect, Failure Modes Effect Analysis,
  Attribute and Variable Data
• Determining Process Capability
• What you need to know to apply SPC
• Building a Process Map, CE Matrix, defining
  Control Charts
• Action Plan Moving Forward

15
Six Sigma Phases and Tools
Define
Analyze
Improve
Control
Measure
Maps and Metrics
Cause and Effect Matrix
What is the process? What is wrong? What are
the goals?
Gage RR Study
Design of Experiments
Capability Analysis
Multi-variable Analysis
Failure Mode and Effects Analysis
SPC
16
Six Sigma Process Steps
Measure

• Phase 1 Process Measurement
• Map Process
• Identify Key Process Input Variables (KPIVs)
  which are Critical To Quality (CTQs) - Use
  surveys, interviews, - Fishbone Chart,
  brainstorming, customer needs mapping, etc
• Establish Measurement System Capability - What
  attribute of CTQ variable can be measured?
• Establish Process Capability Baseline
• Phase 2 Process Analysis
• Complete FMEA - Attack top problems
• Review Capability Data, Pareto Analysis
• Continue to ID KPIVs
• Phase 3 Process Improvement
• Verify and Optimize Critical KPIVs
• Phase 4 Process Control
• Implement Control Plan, SPC
• Verify Long Term Capability

Analyze
Improve
Control
17
Toward an Optimized Process
All Xs 1st Hit List Screened List Found
Critical Xs Controlling Critical Xs
Process Map
CE Matrix and FMEA
Multi-Variable Study
Experimentation
Control Plans/SPC
18
Agenda

• Introduction of Participants
• Introduction to Six Sigma Methodology
• SPC Terminology and Tools of the Trade
• Process Mapping
• Cost and Effect, Failure Modes Effect Analysis,
  Attribute and Variable Data
• Determining Process Capability
• What you need to know to apply SPC
• Building a Process Map, CE Matrix, defining
  Control Charts
• Action Plan Moving Forward

19
Tool 1 - Process Map

• Identify each step of the process
• Goal is to identify sources for variation (SOV)
  at each step
• Data collection opportunity at each decision step
• Identify bottlenecks if cycle time is an issue

Turn Steam On
Paperwork

• Preheating
• Gauge Accuracy

• BOM
• ISO Procedures

At Temp?

• Incoming Temp
• Volume
• Speed

• Load Accuracy
• Humidity
• Mixer Speed

Additive
20
Tool 2 - CE Matrix

• Identify key customer requirements from Process
  Map
• Rank order and assign priority factor (1-10)
• Identify all process steps and materials from
  Process Map
• Evaluate correlation of each input to each output
• Cross multiply correlation values with priority
  factors and add across for each input

21
CE Matrix
Importance to Customer/Project
Key Outputs
Pareto List
Correlation
Key Inputs
22
Tool 3 - Failure Mode and Effects Analysis

• Process Step - What step in the process under
  investigation?
• Potential Failure Mode - In what ways does the
  key input go wrong?
• Potential Failure Effects - What is the impact on
  the output variables?
• SEV- How SEVere is the effect to the customer?
  (1-10 Very)
• Potential Causes - What causes the key input to
  go wrong?
• OCC - How often does the failure mode occur?
  (1-10 Very Likely)
• Current Controls - What existing controls and
  procedures (including inspection and test)
  prevent the cause or the failure mode?
• DET - How well can you detect the custom of
  failure mode? (1-10 Not Likely at all to detect)
• RPN - Risk Priority Number SEV x OCC x DET
• Actions Recommended - To reduce occurrence or
  improve detection - Focus on high RPNs

23
Agenda

• Introduction of Participants
• Introduction to Six Sigma Methodology
• SPC Terminology and Tools of the Trade
• Process Mapping
• Cost and Effect, Failure Modes Effect Analysis,
  Attribute and Variable Data
• Determining Process Capability
• What you need to know to apply SPC
• Building a Process Map, CE Matrix, defining
  Control Charts
• Action Plan Moving Forward

24
Statistical Process Control Concepts

• Two types of Variation
• Common Causes (90)
• Special Causes (10)
• Statistical Process Control will identify the
  presence of special causes
• Allowing them to be identified and eliminated
• If special causes are a frequent occurrence in
  the process then attempts to apply Statistical
  Process Control are futile.

25
Process Capability

• Ensure that the process is capable of
  consistently producing the product to the
  required specification
• To demonstrate capability
• the process must be in a state of statistical
  control special causes must not be present
• the precision of the process must be sufficient
  to keep the proportion of defects below a certain
  minimum level.

26
Process Capability Studies

• Are in Common use in Industry
• They are carried out at various times in the life
  of the process
• Usually on characteristics that are particularly
  critical to customer acceptance
• Example) Characteristics that could impact on
  product liability, compliance with regulation, or
  cost.

27
Capability Studies in Industry use Automation

• Graphical methods based on Normal Probability
  paper can be used without any understanding of
  the theory, but by following a step-by-step
  process
• Plot of results from a non-normal distribution
  will produce a curve, this will alert you to the
  fact that it is a non-normal distribution
• Alternatively a variety of computer programs can
  calculate process capability directly from the
  source data which can be captured automatically
  by the computer.

28
Run Chartand Control Chart

• The Main Difference Between a Run Chart and a
  Control Chart
• The Run Chart Sample Size Is One
• Control Charts Use a Subgroup of Several Samples

29
Choosing theCorrect Control Chart Type
Start
Type of Data
Counting Defects or Defectives?
Are Poisson assumptions satisfied?
Are Binomial assumptions satisfied?
Prop for rejects low?
Know Bad and Good
Area of opportunity consistent from sample to
sample?
Constant Sample Size?
Convert counts input s or rates
Indiv-X, Moving Range
30
Example

• Every hour a subgroup of 4 bottles is taken from
  a bottling process. The weights of the bottles
  are tabulated below. The specification for bottle
  weight is 350-10.
• Find the trial limits for the X-Bar and R charts
  assuming special causes for any out of control
  points
• Find the revised control limits
• Find the process mean and process standard
  deviation
• The proportion of output which is nonconforming
• If the process mean shifts to 345, the
  probability of detecting this shift on the first
  sample drawn after the shift.
• The proportion of output which is nonconforming
  at this value of the process mean.

31
Agenda

• Introduction of Participants
• Introduction to Six Sigma Methodology
• SPC Terminology and Tools of the Trade
• Process Mapping
• Cost and Effect, Failure Modes Effect Analysis,
  Attribute and Variable Data
• Determining Process Capability
• What you need to know to apply SPC
• Building a Process Map, CE Matrix, defining
  Control Charts
• Action Plan Moving Forward

32
Implementing SPC

• Management must ensure the problems encountered
  are tackled objectively
• The causes of variation, and potential problems,
  must be understood
• Taguchi analysis or small group improvement
  techniques
• The process must also be in a state of
  statistical control
• If it is not then it must be brought into a state
  of statistical control

33
Measurement System

• The characteristics to be measured must be
  defined, and the measuring equipment must be
  appropriate and give consistent results.
• The precision of the measuring equipment must be
  sufficient to give at least ten measurement steps
  in the range of process variation.

34
Selecting Subgroups

• Control charts compare variations within a
  subgroup to variations between subgroups
• Subgroups should be chosen so that the
  opportunities for variation within subgroups is
  as small as possible, and due only to common
  causes
• The subgroup is usually small and the individual
  items in a particular subgroup are taken over a
  short period of time
• A subgroup should consist of for four or five
  identical items.
• Sampling frequency should be sufficient to catch
  potential sources of variation over time
• This may be once an hour, twice a shift or some
  other interval
• It is important to consider potential sources of
  variation, such as shift changes in selecting
  sampling frequencies.

35
Initial Study

• Establish control limits
• Carried out over a fairly short time-span
• Must catch the variations over time, such as
  shift changes
• At least thirty samples must be taken to be
  statistically valid.
• Ensure that the process is in control
• If it is not then the special causes should be
  identified and eliminated

36
Statistical Controlunderlies the Practice of
aStatistical Process Control

• If the Process is in a State of Statistical
  Control the Statistical Parameters which describe
  the Process Will Remain Constant Over Time
• The Individual Parts Will Vary, but the
  Underlying Statistics Will Remain Constant
  (Parameters Such As the Mean and Variance Will
  Not Change With Time)
• We Cannot Establish These Parameters Directly.
  The Best We Can Do Is to Take a Sample of Parts
  and Calculate the Descriptive Statistics, Which
  We Use to Infer the Process Parameters
• The Descriptive Statistics Provide Only an
  Estimate, and Our Confidence in the Results Will
  Depend on the Sample Size

37
Agenda

• Introduction of Participants
• Introduction to Six Sigma Methodology
• SPC Terminology and Tools of the Trade
• Process Mapping
• Cost and Effect, Failure Modes Effect Analysis,
  Attribute and Variable Data
• Determining Process Capability
• What you need to know to apply SPC
• Building a Process Map, CE Matrix, defining
  Control Charts
• Action Plan Moving Forward

38
Lets Get Started

• Define
• Process Map
• CE Matrix

39
Why, What, How?

• Why is the project worth doing?
• Why is it important now?
• What are the consequences of not doing this?
• What activities have higher or equal priority?
• How does this fit with business initiatives and
  targets?

40
Project Scope

• What must this project accomplish?
• What resources are available to the team?
• What is out of bounds?
• What constraints must the team work under?

41
Deliverables

• What specific goal must be met? When?
• What milestones are critical?
• What would constitute stretch results?

42
Process Map
43
CE Matrix
44
Appendix
45
Quality Terminology

• Sigma (s) - The distribution about the mean of a
  process - goal is /- 3 s or 6 s total
• Lower Specification Limit (LSL), and an Upper
  Specification Limit (USL).
• CTQ - Critical to Quality - A few measurable
  characteristics of a process where reduced
  variation will have a positive impact on process.
• KPIV - Key Process Input Variables -
• FMEA - Failure Mode Effects Analysis -
• RTY - Rolled Throughput Yield
• DPU - Defects per Unit
• COPQ - Cost of Poor Quality
• DPMO - Defects per Million Opportunities

46
Accurate or Precise?

• The ability of a process to meet specification is
  often referred to as the accuracy of the process.
  This embraces two separate considerations
• the position at which the process is centered,
  which is the true meaning of accuracy
• the dispersion or spread of results, of which the
  standard deviation is a useful measure.

47
Attributeand Variable Data

• Attribute Data (Qualitative)
• Categories
• Yes, No
• Go, No Go
• Pass/Fail
• Variable Data (Quantitative)
• Continuous
• Electronic or Mechanical Testing
• Measurement

48
Acknowledgements

• Allied Signal - Dr. Steve Zinkgraf
• GE - Steve Specker
• Ishikawa, Kaoru - What is Total Quality Control?
  The Japanese Way

49
Want to learn more?

• Contact DST Controls
• [phone]
• [phone]
• [phone] FAX
• Rhayward_at_dstcontrols.com