The Quality Improvement Model

1
TheQualityImprovementModel
Define Process
Select Measures
Collect Interpret Data
Improve Process Capability
IsProcessStable?
Investigate Fix Special Causes
No
Purpose Determine/identify fundamental changes
to the process to improve capability.
Yes
IsProcessCapable?
No
Improve Process Capability
Yes
Use SPC to Maintain Current Process
2
Improving Stable Processes

• Common causes of variation usually cannot be
  reduced by trying to explain the difference
  between high and low points when a process is
  stable.
• All of the data, not just the high and low
  points, are relevant to further analysis.
• Aspects of the system that are always present
  will need to be changed to improve common cause
  variation.
• Stable systems usually require managerial
  attention for improvement. Fundamental changes in
  the system are often required.

Source Brian Joiner
3
Order Entry Process
It has been decided that this process requires
further improvement How do we go about it?
4
Order Entry Errors by Type
Pareto Chart
180
160
140
120
100
80
60
40
20
0
Error in
Incorrect
Priority
Inadequate
Incomplete
Other Errors
Order
Account
Code Not
Billing
Shipping
Amount
Number
Checked
Information
Instructions
Type Error
5
Order Entry ProcessResults of efforts to reduce
errors in order amount.
6
Order Entry Errors by Type
7
Improving Stable Processes - Instrument Measures
Common Problems with Stable Processes
Off Target
Excess Variation
Lower Spec
Upper Spec
Lower Spec
Upper Spec
Or, Both
8
Improving Stable Processes - Instrument Measures
Common Improvement Techniques

• Targeting Processes - Designed Experiments
• Change process variables in a planned manner to
  determine
• the effects of the variables on level and
  variability.
• Design of Experiments Course
• Reducing Variation - Use Process Knowledge
• Break out process steps (Process Map and CE
  information)
• Isolate sources of variability (Exploratory Data
  Analysis or Designed Experiment)
• Separate out common causes (stratification
  Applied Statistical Methods Course)
• Exploratory Data Analysis (Applied Statistical
  Methods Course)
• Multi-Vary Studies in Six Sigma
• Analysis on existing process data
• Done on key variables identified in CE or FMEA

9
Breaking Out Process Steps
Use CE matrix as a place to start data
collection and analysis.
10
Isolating Sources of VariabilityPolymer
Manufacturing Process

• Conduct a study to evaluate measurement
  variability of b.
• Two b measurements are done on each sample.

3.0 2.5 2.0 1.5 1.0 0.5 0
b
Duplicate Measurements on Same Sample
1 2 3 4 5 6
7 8 9
Sample
Possible Actions Concentrate on improving the
measurement process (Previous Module)
11
Separating Out Common CausesShipping Process
Possible Actions Improve process for shipping of
international orders.
12
Hypothesis Are the Labs that Measure Color the
Same?
13
Minitab Comparing Two or More Samples

• Open Mintab Software
• Open Color Round Robin Data in the Mintab
  Datasets Folder and follow the directions in the
  notes below.

Note The ASM course spends more time discussing
Planning, Data Collection, Analysis and
Interpretation from experiments similar to the
one above.
14
Comparing Two or More Labs
Follow the instructions in the notes to perform
the comparisons.
15
Comparing Two or More Labs
Individual 95 CIs For Mean Based on
Pooled StDev Level N
Mean StDev --------------------------------
---- HART 24 -1.7658 0.2235 (---) Kuantan
41 -1.3346 0.1369 (-) TED
37 -1.0568 0.0846
(--)
------------------------------------
-1.75 -1.50 -1.25
-1.00 Pooled StDev 0.1476 Tukey 95
Simultaneous Confidence Intervals All Pairwise
Comparisons among Levels of Lab Individual
confidence level 98.09 Lab HART subtracted
from Lab Lower Center Upper
------------------------------------ Kuantan
0.3408 0.4312 0.5216
(----) TED 0.6169 0.7091 0.8012
(----)
-----------------------------
------- -0.30
0.00 0.30 0.60 Lab Kuantan
subtracted from Lab Lower Center Upper
------------------------------------ TED
0.1981 0.2779 0.3576
(---)
------------------------------------
-0.30 0.00 0.30
0.60
16
Hypotheses Does Film IV, CHDM and DEG Effect
Shrink Stress?
17
Minitab Analyze Relationships Between Two
Continuous Variables

• Open Minitab Software.
• Open Shrink Stress in the Minitab Datasets
  Folder.
• Follow the instructions below to create fitted
  line plots of the data.

18
Analyze Relationships Between Two Continuous
Variables
What conclusions do you make from the analysis?
19
Designed Experiments
Input Variables
Making planned and deliberate changes to the
process input variables to determine the effect
on the output variables.
Properly designed experiments can yield useful
cause-effect information.
20
Designed Experiments - Example
Objective Choose a formulation for a new EMBRACE
product. These 16 conditions were done on a
manufacturing line in Kingsport. Material was
taken to customers and run. Customer fitness for
use measures will aid in determining new EMBRACE
fromulation.
21
Designed Experiments - Example
22
Review

• Improving Stable Processes takes intervention
• All of the data is important, not just the
  "Special" Causes
• Exploratory Data Analysis (EDA) on historical
  data
• Use the CE Matrix and the FMEA to determine
  where to start
• Look for trends
• Look for differences among qualitative variables
  (instruments, etc.)
• Information will be used in control plan
• Variance Component Breakdowns (Test and Process)
• Designed Experiments on Key Inputs
• Key Inputs determined by CE, FMEA and EDA
• Information will be used in control plan

23
Exercise
1.) Your Catapult Team should complete page 12 of
the Catapult Process handout. Limit
yourselves to 30 minutes for this exercise.