QUALITY MANAGEMENT

1
QUALITY MANAGEMENT

• Defining Quality
• Total Quality Control
• The Costs Of Quality
• Statistical Process Control

2
DEFINING QUALITY

• What is product "quality"?
• Those factors that influence customer
  satisfaction
• Design quality -- is the product designed to
  satisfy customer needs?
• Conformance quality -- is the product
  manufactured to design specs?

3
TOTAL QUALITY CONTROL

• Design quality control -- ensure product is both
  marketable and manufacturable
• Product/service control -- ensure facility
  achieves conformance quality
• Process capability studies is proposed process
  capable?
• Statistical process control potential problem
  detection after capable process is selected
• Incoming material quality control -- ensure
  suppliers achieve conformance quality
• Acceptance sampling detect bad shipments

4
THE COSTS OF QUALITY

• Isnt TQC expensive? Not as expensive as not
  doing it
• This is illustrated by identifying the costs of
  quality at a facility
• Costs of not doing TQC
• Costs of internal failure (scrap)
• Costs of external failure (warranties)
• Costs of doing TQC wrong
• Costs of inspection
• Costs of doing TQC right
• Costs of prevention (TQC system)

5
REWARDING QUALITY

• Deming Prize
• Awarded by Japan to firms achieving high quality
  through SPC
• Baldridge Award
• Awarded by US government to firms achieving high
  quality as demonstrated by customer satisfaction
• 6 awards maximum per year
• 75-page application

6
CERTIFYING QUALITY

• ISO 9000
• Series of standards for a quality management
  system
• Say what you do, do what you say
• Latest version more focused on results
• QS 9000
• ISO 9000-based series of standards customized for
  the auto industry
• ISO 14000
• Series of standards for an environmental
  management system

7
ATTRIBUTE VERSUS VARIABLE QUALITY CHARACTERISTICS

• To manage, control, and improve quality, we must
  measure it
• Quality characteristics (QCs)_are what we
  measure
• Dimensions
• Other chemical, physical properties
• Defects per unit
• Quality characteristics may be
• Variable (continuous) data
• Attribute (discrete) data

8
PROCESS CAPABILITY INDICES

• Used to ensure that a process is capable of
  making the part to spec when everything is
  working properly
• µ, s -- Mean, standard deviation of QC for
  stable process
• UNL, LNL -- Upper, lower natural limits
• The range of variation we expect to see when
  things are working right
• USL, LSL -- Upper, lower spec limits
• The range of variation we can tolerate for the
  quality characteristic

9
PROCESS CAPABILITY INDICES

• Natural limits (quantitatively) defined
• A normally distributed QC from a stable process
  will fall within the natural limits 99.73 of the
  time
• This implies that the natural limits are
•  UNL µ 3s
•  LNL µ - 3s
• The basic process capability index is thus

• Generally OK if Cp gt 1.33

10
PROCESS CAPABILITY INDICES

• Suppose we are cutting 6 metal brackets
• LSL, USL 5.95, 6.05
• m 6
• s 0.01

11
STATISTICAL PROCESS CONTROL

• Once we know our process is capable, we use SPC
  to detect potential problems
• This means distinguishing between
• Unassignable error -- randomness due to inherent
  variability in the process
• The process is in control
• Assignable error -- variability resulting from a
  change in the process
• The process is out of control

12
CONTROL LIMITS FOR PROBLEM DETECTION

• A control chart is a graphical aid to help the
  worker make this distinction
• UCL -- upper control limit for Q.C.
• LCL-- lower control limit for Q.C.

13
CONTROL LIMITS FOR PROBLEM DETECTION

• If an observation falls inside the limits
• Conclude process is in control
• If an observation falls outside the limits
• Conclude process is out of control

14
MISTAKES TO AVOID

• Need to set control limits to minimize two types
  of mistakes
• Type I -- conclude we're out of control when
  we're in control
• Type II -- conclude we're in control when we're
  out of control
• Probability of each type of mistake
• Type I -- a Type II-- b
• H0 the process is in control

15
SETTING CONTROL LIMITS

• Focus on minimizing Type I error first
• If control limits set to /- 3 standard
  deviations of the statistic, probability of a
  Type I error lt 1
• Assumes normal distribution
• What about Type II error?
•  Increase sample size

16
P CHARTS FOR ATTRIBUTES

• Tracks "go/no-go" quality characteristics
•  n -- Sample size
• m Number of samples collected to date
•  Xi -- Number of nonconformities in ith sample
•  pi Fraction nonconforming in the ith sample
• Xi is a binomial random variable
• Example Leaking juice cans
• n 100
• m 10

17
P CHARTS FOR ATTRIBUTES

• Estimate the mean fraction nonconforming by
  calculating p-bar from all data

• Estimate the standard deviation of the fraction
  nonconforming in the samples

18
P CHARTS FOR ATTRIBUTES

• Set control limits 3 standard deviations above
  and below the mean

19
X-BAR AND R CHARTS FOR VARIABLES

• A normal distribution has two parameters
• Mean (Location)
• Variance (Dispersion)
• A control chart is needed to track each parameter
• If either one or both is out of control, the
  process is out of control
• X-bar chart tracks mean
• R chart tracks dispersion

20
X-BAR AND R CHARTS

• Tracks continuously variable quality
  characteristics
•  n -- Sample size
• m Number of samples collected to date
• X-bari Mean of ith sample
• Ri Range of ith sample

• X-bari are approximately normally distributed
• Example Bursting strength (in psi) of 1-liter
  soda bottles
• n 4
• m 10

21
X-BAR AND R CHARTS

• Tracks continuously variable quality
  characteristics
•  n -- Sample size (4)
• m Number of samples collected to date (10)
• X-bari Mean of ith sample
• Ri Range of ith sample

22
X-BAR AND R CHARTS

• Then the grand mean (X-bar-bar) and the mean
  range (R-bar) are

23
X-BAR CHART

• The control limits for the charts are in practice
  obtained using a table of control chart constants

24
R CHART

• The control limits for the charts are in practice
  obtained using a table of control chart constants

25
INCOMING MATERIAL QUALITY CONTROL

• Objective ensure vendor delivers quality
  supplies
• Two ways to reach objective
• Inspect vendor's shipments
• Acceptance" sampling
•  
• Have vendor performing TQC
• Vendor supplies customer with relevant control
  charts
• Customer certifies vendor in TQC

26
ACCEPTANCE SAMPLING

• Basic idea
• Inspect a random sample of each lot
• Classify each item as ok/not ok
• Conclude entire lot is either
• Ok -- accept it
• Not ok -- reject it and/or sort it