SPC Training

SPC : INTRODUCTION

Statistical Process Control has now been incorporated by

organizations around the world as a primary tool to improve product
quality by reducing process variation.

SPC is a technique that employs statistical tools for controlling and

improving processes. SPC is an important ingredient in continuous
process improvement strategies. It uses simple statistical means to
control, monitor, and improve processes.

1 OF 28
 Variation:  The Heart of SPC

• Variation exists in everything

• When we’re manufacturing products with customers that expect and

demand high quality and consistency in our goods, variation can
become a big problem.

• Too much variation leads to rework, scrap, or customer problems.

• A perfect process would be one with no variation. They don’t exist.

• As the variation in our processes is reduced, the output of our

processes will be improved.

• That’s our goal with SPC-to reduce the variation in our processes and
then monitor the process to make sure the variation doesn’t increase.

2 OF 28
Continuous Improvement Steps

Define and Prioritize Measure and Collect Analyze Data and

Problems Data Identify Root Cause

Pareto Chart Measurement

Fishbone Diagram
System Analysis
Process Flow Chart
Trend Chart Histogram
Value-added Flow 5-Why Analysis
Statistical Process
Chart
Control - SPC Regression
Project Priority Analysis
Calculator
Six Sigma Design of
Failure Mode & Conversion Table Experiments
Effects Analysis Statistics
FMEA Handbook

Improve Processes Control Processes

Brainstorming Control Plan

System Diagrams Statistical Process

Control - SPC
Error-Proofing
Corrective Action CHECK Process
Matrix Failure Mode &
CHECK Process Effects Analysis
FMEA
5S
3 OF28
 STATISTICAL PROCESS CONTROL (SPC) TOOLS

Among the many tools for quality improvement, the following are the most
commonly used tools of SPC:

Histograms
Cause-and-Effect diagrams
Pareto Diagrams
Control Charts
Scatter or Correlation Diagrams
Run Charts
Process Flow Diagrams

Figure 1 shows the seven basic tools of statistical process control,

sometimes known as the "magnificent 7."

4 OF 28
5 OF 28
 PARETO CHART EXAMPLE

Pareto Chart of CLAIMS CODE

140 100
120
NUMBER OF CLAIMS

80
100

Percent
80 60

60 40
40
20
20
0 0
CLAIMS CODE r or k r r r
rro r oc ro ro he
E Er S t Er Er
O t
n s g
tio a le
rom rin oom
duc S F e e R
o ts in k
Pr Pa
r g oc
d En S t
Ba
NUMBER OF CLAIMS 68 29 12 12 12 6
Percent 48.9 20.9 8.6 8.6 8.6 4.3
Cum % 48.9 69.8 78.4 87.1 95.7 100.0

6 OF 28
 Cause & Effect Diagram: Interface Defects

ENG. DOC TOOLING OPERATOR

TRIMS SPEC. TOO INADEQUATE

WORN OUT
LONG TRAINING

ERRORS IN LACK OF PREVENTIVE LACK OF SELF

DESIGN/DOC. MAINTENANCE INSPECTION

LACK OF DETAILED IMPROPER

NOT FOLLOWING
ROUTING & TOOLING &
PROCEDURE
INSTRUCTIONS SET-UP

DIRTY

CONTACTS
TOO
HIGH/LOW

HEAT FROM
NOT TO MFG. INADEQUATE
SOLDERING CAUSES
SPEC INCOMING INSPECTION
CONTACTS TO MOVE

LACK OF PROPER
IN_PROCESS
BENDING/ INSPECTION TOOLS
TRIMMING

OTHER PROCESSES MATERIAL INSPECTION

7 OF 28
 BASIC STATISTICS

• MEAN

• RANGE

• STANDARD DEVIATION

• PPM

8 OF 288
 Set A Set B
Measure 2, 2, 3, 5, 5, 7, 8 2, 3, 3, 4, 6, 7
The Mean Adding the numbers up gives: Adding the numbers up gives:
To find the mean, you 2 + 2 + 3 + 5 + 5 + 7 + 8 = 32 2 + 3 + 3 + 4 + 6 + 7 = 25
need to add up all the There are 7 values, so you divide There are 6 values, so you divide
data, and then divide the total by 7:    32 ÷ 7 = 4.57... the total by 6:    25 ÷ 6 = 4.166...
this total by the number So the mean is 4.57 (2 d.p.) So the mean is 4.17 (2 d.p.)
of values in the data.

The Median The numbers in order: The numbers in order:

To find the median, you 2 , 2 , 3 , (5) , 5 , 7 , 8 2 , 3 , (3 , 4) , 6 , 7
need to put the values The middle value is marked in This time there are two values in
in order, then find the brackets, and it is 5. the middle. They have been put
middle value. If there are in brackets. The median is found
two values in the middle So the median is 5 by calculating the mean of these
then you find the mean two values:    (3 + 4) ÷ 2 = 3.5
of these two values. So the median is 3.5

The Mode The data values: The data values:

The mode is the value 2,2,3,5,5,7,8 2,3,3,4,6,7
which appears the most The values which appear most This time there is only one value
often in the data. It is often are 2 and 5. They both which appears most often - the
possible to have more appear more time than any number 3. It appears more times
than one mode if there of the other data values. than any of the other data values.
is more than one value
which appears the most. So the modes are 2 and 5 So the mode is 3

The Range The data values: The data values:

To find the range, you 2,2,3,5,5,7,8 2,3,3,4,6,7
first need to find the The lowest value is 2 and the The lowest value is 2 and the
lowest and highest values highest value is 8. Subtracting highest value is 7. Subtracting
in the data. The range is the lowest from the highest the lowest from the highest
found by subtracting the gives:    8 - 2 = 6 gives:    7 - 2 = 5
lowest value from the So the range is 6 So the range is 5
highest value.

9 OF 28
standard deviation is a measure of the variability of a data set. A low
standard deviation indicates that the data points tend to be very close to the
same value (the mean), while high standard deviation indicates that the data
are spread out over a large range of values.

The variance of the population mean is denoted by   

                  

The standard deviation is the most important measure of variability. It is

the square root of the variance.

                         

10 OF 28
 Implementing Statistical Process Control
1. Determine Measurement Method

Statistical Process Control is based on the analysis of data, so the first

step is to decide what data to collect. There are two categories of
control chart distinguished by the type of data used: Variable or
Attribute. Variable data comes from measurements on a continuous
scale, such as: temperature, time, distance, weight. Attribute data is
based on upon discrete distinctions such as good/bad, percentage
defective, or number defective per hundred.

2. Qualify the Measurement System

A critical but often overlooked step in the process is to qualify the

measurement system. No measurement system is without
measurement error. If that error exceeds an acceptable level, the data
cannot be acted upon reliably.
GAUGE R & R (Reliability & Repeatability)
11 OF 28
 Gauge R & R studies

The basic issue for the study is the behavior of gauges in a particular
environment with respect to:

Repeatability : Does it produce same result?

Reproducibility : Is it dependent on skills of Operator?
Stability : Does it change over time?

12 OF 28
 GAUGE R&R FOR AM-40 ANGLE MEASURENT TOOL

Trial Operator A Operator B

  Reading 1 Reading 2 X bar Reading 1 Reading 2 X bar R bar Xbar*

                 

1 132.8 132.7 132.75 133.2 132.8 133 0.5 132.875

2 130.2 129.8 130 130.3 130.6 130.45 0.8 130.225

3 132.5 131.9 132.2 132.6 132.7 132.65 0.8 132.425

4 127.9 128.4 128.15 128.5 128.5 128.5 0.6 128.325

5 133.1 133.5 133.3 133.5 133.6 133.55 0.5 133.425

6 131.8 131.4 131.6 132.2 132.2 132.2 0.8 131.9

7 128.6 129.3 128.95 129.3 129.2 129.25 0.7 129.1

8 133.4 133.2 133.3 133.3 133.2 133.25 0.2 133.275

9 136.7 136.4 136.55 136.8 136.7 136.75 0.4 136.65

10 130.2 130.1 130.15 130.1 130.1 130.1 0.1 130.125

      131.695     131.97    

13 OF 28
 Gage R&R Study - XBar/R Method

Study Var %Study Var
Source StdDev (SD) (6 * SD) (%SV)
Total Gage R&R 0.29179 1.7507 11.21(< 20% GOOD)
(Up to 30% Acceptable)
(> 30% Not Acceptable)
Repeatability 0.23493 1.4096 9.03
Reproducibility 0.17305 1.0383 6.65
Part-To-Part 2.58648 15.5189 99.37
Total Variation 2.60288 15.6173 100.00

Number of Distinct Categories = 12

14 OF 28
 3. Initiate Data Collection and SPC Charting

Develop a sampling plan to collect data (subgroups) in a random fashion at a

determined frequency. Be sure to train the data collectors in proper measurement
and charting techniques. Establish subgroups following a rational subgrouping
strategy so that process variation is captured BETWEEN subgroups rather than
WITHIN subgroups. If process variation (e.g. from two different shifts) is captured
within one subgroup, the resulting control limits will be wider, and the chart will be
insensitive to process shifts.
The type of chart used will be dependent upon the type of data collected as well as
the sample size, as shown by the table below.

15 OF 28
 4. Assess Control.

After establishing control limits, the next step is to assess whether or not the process is
in control (statistically stable over time). This determination is made by observing the
plot point patterns.
When an out-of-control condition occurs, the points should be circled on the chart, and
the reaction plan should be followed.

16 OF 28
zone A = mean +/- 3 sigma
zone B = mean +/- 2 sigma
zone C = mean +/- 1 sigma
17 OF 28
18 OF 28
5. Analyze Data to Identify Root Cause and Correct

You ARE OUT OF CONTROL IF . . .

A single point falls outside the 3 Sigma limit, i.e. beyond Zone 3

Two out of three successive points fall in Zone 3 or beyond, the odd point may be anywhere

Four out of Five successive points fall in Zone 2 or beyond, the odd point may be anywhere

6. Design and Implement Actions to Improve Process Capability

19 OF 28
 What is Process Capability?
Process capability compares the output of an in-control process to the
specification limits by using capability indices. The comparison is made by
forming the ratio of the spread between the process specifications (the
specification "width") to the spread of the process values, as measured by 6
process standard deviation units (the process "width").
A capable process is one where almost all the measurements fall inside
the specification limits. This can be represented pictorially by the plot
below:

20 OF 28
 % of non
% of products in Non conformance
Range around Mean conforming
conformance out of a million
products

-1σ  to  + 1σ 68.46 31.54 315400

       

-2σ  to  + 2σ 95.46 4.54 45400

       

-3σ  to  + 3σ 99.73 0.27 2700

       

-4σ  to  +4σ 99.9937 0.0063 63

       

-5σ  to  + 5σ 99.999943 0.000057 0.57

       

-6σ to  + 6σ 99.9999998 0.00000002 0.002

21 OF 28
7. Calculate Cp and Cpk and Compare to Benchmark

22 OF 28
 Process Capability : Bending 134 deg. angle

Process Capability of Readings

LSL USL
P rocess Data W ithin
LS L 132 O verall
Target *
USL 136 P otential (Within) C apability
S ample M ean 131.843 Cp 0.39
S ample N 40 C PL -0.03
S tDev (Within) 1.70258 C PU 0.81
S tDev (O v erall) 2.36273 C pk -0.03
O v erall C apability
Pp 0.28
PPL -0.02
PPU 0.59
P pk -0.02
C pm *

128 130 132 134 136

O bserv ed P erformance Exp. Within P erformance E xp. O v erall P erformance
P P M < LSL 475000.00 PPM < LSL 536852.20 P P M < LS L 526573.85
P P M > U S L 100000.00 PPM > USL 7305.52 PPM > USL 39236.56
P P M Total 575000.00 PPM Total 544157.72 P P M Total 565810.41

NOTE: Without Tweaking (reworking), more than 50% of the parts would be rejecyed. 23 OF 28
 Process Capability : Marker Cutting

LSL USL
P rocess D ata W ithin
LS L 0.94 Ov erall
Target *
USL 1.06 P otential (Within) C apability
S ample M ean 0.9897 Cp 3.51
S ample N 20 C P L 2.91
S tDev (Within) 0.00569242 C P U 4.12
S tDev (O v erall) 0.00490005 C pk 2.91
O v erall C apability
Pp 4.08
PPL 3.38
PPU 4.78
P pk 3.38
C pm *

0.944 0.960 0.976 0.992 1.008 1.024 1.040 1.056

O bserv ed P erformance E xp. Within P erformance E xp. O v erall P erformance

P P M < LS L 0.00 P P M < LS L 0.00 P P M < LS L 0.00
P P M > U S L 0.00 P P M > U S L 0.00 P P M > U S L 0.00
P P M Total 0.00 P P M Total 0.00 P P M Total 0.00

24 OF 28
 EXAMPLE OF Cpk IMPROVEMENT

Process Capability : Sleeve Cutting

LSL USL
P rocess D ata Within
LS L 4.9 Overall
Target *
USL 5.02 P otential (Within) C apability
S ample M ean 5.0134 Cp 2.18
S ample N 20 C P L 4.11
S tDev (Within) 0.00919186 C P U 0.24
S tDev (O v erall) 0.00914445 C pk 0.24
O v erall C apability
Pp 2.19
PPL 4.13
PPU 0.24
P pk 0.24
C pm * Process Capability of After CF

LSL USL
P rocess D ata Within
4.90 4.92 4.94 4.96 4.98 5.00 5.02 LS L 1.44 Ov erall
Target *
O bserv ed P erformance E xp. Within P erformance E xp. O v erall P erformance USL 1.56 P otential (Within) C apability
P P M < LS L 0.00 PPM < LS L 0.00 P P M < LS L 0.00 S ample M ean 1.5109 Cp 7.52
P P M > U S L 300000.00 PPM > U S L 236370.51 P P M > U S L 235224.44 S ample N 30 C P L 8.89
P P M Total 300000.00 PPM Total 236370.51 P P M Total 235224.44 S tDev (Within) 0.00265957 C P U 6.15
S tDev (O v erall) 0.00292846 C pk 6.15
O v erall C apability
Pp 6.83
PPL 8.07
PPU 5.59
P pk 5.59
C pm *

1.440 1.456 1.472 1.488 1.504 1.520 1.536 1.552

O bserv ed P erformance Exp. Within P erformance Exp. O v erall P erformance

P P M < LS L 0.00 P P M < LS L 0.00 P P M < LS L 0.00
P P M > U S L 0.00 P P M > U SL 0.00 P P M > U S L 0.00
P P M Total 0.00 P P M Total 0.00 P P M Total 0.00

25 OF 28
 Process Capability Sixpack of Imterface 7203-1541-021
I Chart Capability Histogram
LSL USL
0.0090 UCL=0.008996
S pecifications
Individual Value

LS L 0.000
_
0.0075 X=0.0075 U S L 0.015

0.0060 LCL=0.006004 0 2 4 6 8 0 2 4
00 .00 . 00 .00 . 00 .01 . 01 .01
1 3 5 7 9 11 13 15 17 19 21 23 25 0. 0 0 0 0 0 0 0

Normal Prob Plot

Moving Range Chart
0.002 A D: 1.493, P : < 0.005
UCL=0.001838
Moving Range

0.001
__
MR=0.000563

0.000 LCL=0
1 3 5 7 9 11 13 15 17 19 21 23 25 0.006 0.007 0.008 0.009

Last 25 Observations Capability Plot

0.0084 Within
Within O v erall
S tDev 0.00049867 S tDev 0.000478714
Values

0.0078
Cp 5.01 Pp 5.22
O v erall
C pk 5.01 P pk 5.22
0.0072 C pm *
5 10 15 20 25 S pecs
Observation

26 OF 28
CONCLUSION:

SPC = Statistical Process Control

SPC = Data Driven Process, No guess
work
SPC = Control & Minimize Variations in
Process
SPC = Identifies Capabilities and
Limitations of Process
SPC = Helps Identifying areas for
Improvement in Process
SPC = Higher Customer Satisfaction &
Improved Profits $$$

27 OF 28