STATISTICAL PROCESS CONTROL

The primary goals of SPC

• Minimize production costs.
• Attain a consistency of products and services that
will meet production specifications and customer
expectations.
• Create opportunities for all members of the
organization to contribute to quality improvement.
• Help both management and employees make
economically sound decision about actions affecting
the process.
• Seven basic tools that have been used
successfully in food industries for decades:
§ Flow chart
§ Cause and effect diagram
§ Control chart
§ Histogram
§ Check sheet
§ Pareto chart
§ Scatter diagram
 What is a Process Flow Chart?

• 1. A Process Flowchart is a graphic

representation of all the major steps of a
process.
It can help you:
» a.
Understand the complete process.
» b.
Identify the critical stages of a process.
» c.
Locate problem areas.
» d.
Show relationships between different steps in
a process.
» e. To identify and eliminate Waste

4
 Why use a Flow Chart?
Process Flowcharts or Process Mapping helps
you understand and improve a process by
enabling you to:
Look for opportunities for improvement (gap
between reality and ideal).
Identify where defects are found (and where
they are introduced).
Identify where measurements need to be put in
place.
See where rework and bottlenecks occur.
Elimination of Non Value Adding activities (Waste)

5
 Flowchart/Process Map levels
There are three major form of Process Map. These are:

Macro: High Level Overview; Entire Process; Links

with other Processes (the Big Picture)

Micro: Operational Level; Practical Links between

Inputs & Outputs; Operational Measurements
& Controls

Hyper-Micro: Design Level; Details into Operator

Methods; Material or Information (Input)
Specification (Accuracy & Precision);
System Functionality

6
 Considerations when Flowcharting/Process Mapping

Remember S-I-P-O-C (Supplier-Inputs-Process-Outputs-

Customer). Start with C, then O, then S, then I, then P
• What level and details are required?
• What are the Measurable outputs?
• Space, Time allowances (best time, day of the week to do
it is ?)
• Group Dynamics: Leader, Facilitator, Size of Group
• If available and appropriate, use Electronic versions

7
 How to do process flow chart

As a team, define:
• process boundaries
where does the process start
what does the process include/not include
where does the process end

• level of detail;
• key process inputs (supplier) and outputs
(customer requirements);
• identify and sequence process steps;
• discuss and analyse the process;
• agree next review.

8
 CAUSE & EFFECT DIAGRAM
(Fishbone or Ishikawa Diagram)

TRNC
 CAUSE & EFFECT DIAGRAM
• The goal of problem solving is to identify
the causes of problems in order to correct
them.
• The cause-and-effect diagram is an
important tool in this task;
– it assists the generation of ideas for problem
causes and, in turn, serves as a basis for
solution finding.
 Preview, contd...
• The cause-and-effect diagram was introduced
in Japan by Kaoru Ishikawa.
• It is referred to as Ishikawa diagram due to its
originator or as a fish-bone diagram because
of its structure.
 Why use it?
Cause-and-effect diagrams are tools in
allowing a team to identify, explore, and
graphically display, in increasing detail, all of
the possible causes related to a problem or
condition to discover its root cause(s).
 What does it do?
• Enables teams to focus on the content of the
problem, not on the history of the problem
• Creates a snapshot of the collective
knowledge and consensus of a team around a
problem.
– This builds support for the resulting solutions
• Focuses the team on causes, not symptoms
 How do I do it?
1. Generate the causes needed to build a
Cause & Effect Diagram. Choose one method:
– brainstorming
– check sheets based on data collected by team
members before the meeting
 How do I do it? contd...

2. Construct the Cause & Effect diagram

– Place the problem statement in a box on the
right hand side of the writing surface.
– Draw major cause categories or steps in the
production or service process. Connect them
to the “backbone” of the fishbone chart.
 Major types of causes
• Production Process
– machines (equipment)
– methods (how work is done)
– materials (components or raw materials)
– people (the human element)
• Service Process
– policies (Higher level decision rules)
– procedures (steps in a task)
– plant (equipment and space)
– people
 Major types of causes, contd...
In both types of processes, environment
(buildings, logistics, and space) and
measurement (calibration and data collection)
are also frequently used.
Remember: There is no perfect set or number
of categories of causes. The user must make
them fit the problem at hand.
 Fishbone Diagram
Measurement Human Machines

Faulty testing equipment Poor supervision Out of adjustment

Incorrect specifications Lack of concentration Tooling problems

Improper methods Inadequate training Old / worn

Quality
Inaccurate Problem
temperature
Defective from vendor Poor process design
control
Ineffective quality
Not to specifications management
Dust and Dirt
Material- Deficiencies
handling problems in product design

Environment Materials Process

Problems with Airline Customer Service
Incorrect delivery problem
 Control Charts

Ø All processes have variation. There are two types:

Ø Common Causes or natural variation
Ø Special Causes or unnatural variation
– Common-cause variation = normal,
random variation
– Special-cause variation = Changes in the pattern of data that can
be assigned to a specific cause

Ø Control charts use upper and lower control limits that

reflect the natural limits of variation in the process.
Ø Based on statistical principals, control charts allow for
the identification of unnatural (nonrandom) patterns
which then identify opportunities for process
improvement.
 Purpose of Control Charts

Ø To improve the process:

§ Attentive use of control charts can identify assignable
causes
§ Control charts can only detect when the process is out-
of-control, not why the process is out of control

Ø To control the process:

§ Process control prevents bad product from being made
§ Determine process capability and stability

Ø What happens after an out-of-control situation

occurs is the core of a successful SPC program.
 Benefits of Control Charting

Ø Proven technique for improving productivity

Ø Effective in defect prevention
Ø Provides diagnostic information
Ø Provides information about process capability

Note: The biggest risk with control charting is making

unnecessary process adjustments. Need to establish
strong rules when implementing Control Charting.
 Components of a Control Chart
Upper
Control Center
Limit Line

Lower
Control
Limit

Control Limits are the Voice of the Process

Spec Limits are the Voice of the Customer

Spec Limits DO NOT belong on Control Charts!

 General Model
• General model for a control chart
• Let W be a sample statistic
– Mean of W is µW
– Standard deviation of W is sW
• CL, UCL, and LCL

• k is the distance from the CL

• Common choice is k=3
• First proposed by Shewhart
 Statistical Basics

How the Control Chart Limits are

Calculated:

UCL = µ + ks
Center Line = µ
LCL = µ - ks
Where :
µ = the sample Mean
s = the sample standard deviation
k = some distance of the control limits
from the center line (usually 3)
 Rational Subgrouping - Definition
Ø A group of units produced under the same set of conditions.

Ø Subgroups (or rational subgroups) are meant to represent a "snapshot" of the

process. Therefore, they must be taken close together in time but still be
independent of each other.

Ø Example:
§ A die cut machine produces 100 parts per hour. The operator measures
five randomly selected parts at the beginning of every hour. Each
sample of five parts is a subgroup.
 Control Chart Roadmap

Ø Select the appropriate variable to control chart – should be a

critical variable, proven via your analysis
Ø Select Type of Control Chart
Ø Establish basis for rational sub-grouping
Ø Determine sample size
Ø Determine measurement method/criteria
Ø Perform initial capability study to establish control limits
Ø Arrange for charting the data
Ø Prepare procedures & train personnel – provide specific instruction
for when to adjust the process
Ø Implement & monitor
Ø Identify special causes through pattern recognition
Selecting the Right Type of Control Chart

29
 I-MR Chart
Ø An I-MR chart plots individual observations (I chart) and moving
ranges (MR chart) over time for variables data.

Ø Use this combination chart to monitor process when it is difficult

or impossible to group measurements into subgroups.
§ This occurs when measurements are expensive, production
volume is low, or products have a long cycle time.

Ø When data are collected as individual observations, you cannot

calculate the standard deviation for each subgroup. The moving
range is an alternative way to calculate process variation by
computing the ranges of two or more consecutive observations
 Xbar-R Chart
Ø Most popular type of control chart

Ø An Xbar-R chart plots the process mean (Xbar chart) and process
range (R chart) over time for variables data collected in
subgroups.

Ø You can use the Xbar-R chart when your subgroup size is 8 or
less. (Note: Use the Xbar-S chart when your subgroup size is 9 or
more.)

Ø Example: A plastics manufacturer wants to determine whether the production process

for a new product is in control. An engineer samples 5 units every hour for 20 hours
and assess the strength of the plastic.
 Xbar-S Chart
Ø An Xbar-S chart plots the process mean (Xbar chart) and
process standard deviation (S chart) over time for variables data
in subgroups.

Ø Use the Xbar-S chart when your subgroup size is 9 or more. You
can use the Xbar-R chart when your subgroup size is 8 or less.

Ø Example: A paint manufacturer wants to assess the stability of

the process used to fill paint cans. An engineer collected
subgroups of 10 cans every hour and used an Xbar-S chart to
monitor the mean and variation of the filled paint cans.
 P Chart
Ø The P chart plots the proportion of nonconforming units (also called
defectives).

Ø For example, you can use a P chart to monitor the following:

§ The proportion of flights that depart late
§ The proportion of bicycle tires that are flat
§ The proportion of product packages that are leaky

Ø While a unit may have many quality characteristics that can be evaluated, it
is always considered as either conforming or nonconforming

Ø The P chart is the most widely used attribute control chart.

 U Chart
Ø The U chart plots the number of defects (also called nonconformities) per unit. It is
possible for a unit to have one or more defects but still be acceptable in function and
performance.

Ø For example, you can use a U chart to monitor the following:

§ The number of tears and pulls per 50 running feet of carpet
§ The number of leaky packages per 50 sachets

Ø You can inspect for one type of defect such as leaky packages. You can also inspect for
several defects together such as leaky packages and blurry labels.
 When to Take Action?
• One point plots outside the Action Limits.
• Eight consecutive points plot on one side of
the Center Line
• Six points plots steadily increasing or
decreasing
• When an unusual or nonrandom pattern is
observed
 When to Take Action?
4

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

-1

-2

-3

-4
Month
 When to Take Action?
4

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

-1

-2

-3

-4
Month
 When to Take Action?
4

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

-1

-2

-3

-4
Month
 Histograms
• Bar chart
• Used to graphically represent groups of data
• Histograms are used to show the different
frequencies in a process
• They are useful in identifying trends and
relationships that can lead to quality improvements.
 Constructing a Histogram
From a set of data compute
• sum
• mean (x)
• Max
• Min
• Range (max-min)
 Constructing a Histogram
• Use range to estimate beginning and end
• Calculate the width of each column by
dividing the range by the number of columns

Range
= Width

# of Columns
 Acme Pizza Example
• Let’s say the owner wants a distribution of
Acme’s Thursday Night Sales
Data Set from last Thursday(slices)
0212241312122434143223212212214221212212121212
1212221212112223142232221232242244122232212242
12421721223121121222122121222424
 Acme Pizza Example
Mean = 2.032258
Max = 7
Min = 0
Range = 7

Question
For 7 columns what would the width be?

Range/Columns=7/7=1 slice
 Acme Pizza Example
Histogram

70 65

60

50

40 33
30
# times ordered

20
12
8
10
0 0 1
0
1 2 3 4 5 6 7
Slices of Pizza
 Constructing a Histogram
How is this helpful to Acme?
• 2 slices of pizza most common order placed
• Distribution of sales useful for forecasting next
Thursday’s late night demand

If you were an Acme manager how could you apply

this information?
 CHECK SHEETS
• A check sheet is a table or a form used to systematically
register data as it is collected.
• Check sheets help organize data by category. They show how
many times each particular value occurs, and their
information is increasingly helpful as more data are collected.
• Main applications of check sheet include registering how
often different problems occur and registering the frequency
of incidents that are believed to cause problems.
• The function of a check sheet is to present information in an
efficient graphical format.
Product quality inspection check sheet
for production management
 - Pareto Analysis (Pareto Charts)
The Pareto Principle:
q Also referred to as the 80-20 rule.
q States that 80% percent of the problems or effects come from
20% of the causes.
q Focuses on identifying the ‘vital few’ from the ‘trivial many’.
q Helps focusing on what really matters.
 - Pareto Analysis
The Pareto Principle:
q The exact percentages may vary in each situation.
q However, most of the activity is caused by relatively few of its
factors.

20%

80%
 - Pareto Analysis
Examples:
q 20% of car drivers cause 80% of the accidents.
q 20% percent of workers do 80% of the work.
q 20% of a company’s clients are responsible for 80% of its revenue
q 20% of the time spent on a task leads to 80% of the results.
q 80% of the customer complaints come from 20% of customers.
q 80% of the wealth belongs to 20% of the population.

20% 80% 20% 80%

Causes Effects Efforts Results
 - Pareto Analysis
The Pareto Principle:
q Used when we have many problems or projects
and we want to focus on the most significant ones.
• Helps prioritize the improvement opportunities
that bring the most value to the business.
• Allows to reach a consensus about what needs to
be addressed first.

q Used during improvement projects to focus on

the causes that contribute most to a particular effect.
 - Pareto Analysis
The Pareto Chart:
q A frequency bar chart.
q The most frequent activities are placed in order from left to
right.
q Normally plots the frequencies of categorical data:
• Such as defects and errors.
q The horizontal axis represents the types of activities:
• Such as issues, problems or causes.
q The vertical axis represents the frequencies
of those activities.
 - Pareto Analysis
The Pareto Chart:
q By arranging the bars from largest to smallest, the vital few
activities can be easily addressed to have greater attention.
q If there are a lot of small or infrequent factors, consider adding
them together into an “other” category.
100%

You may optionally have

40
q 80%

a cumulative line above the 30

60%

bars so that the cumulative 20 40%

percentages can be read from 10 20%

the right vertical axis.
0 0%
E C A D B F
 - Pareto Analysis
The Pareto Chart:
q If the resulted Pareto chart clearly illustrates a Pareto pattern,
this suggests that only few causes account for about 80% of the
problem.
q This means that there is a Pareto effect.
q If no Pareto pattern is found, we cannot say that some factors
are more important than others.
 - Pareto Analysis
How to Construct a Pareto Chart:
q Define the problem.
q Identify the possible causes of the problem
(using brainstorming or similar technique).
q Collect then record the data.
q Calculate the frequencies of the identified causes.
q Draw a vertical bar for each cause or cause group.
q Sort them by frequency in descending order.
q Calculate then draw the cumulative percentage line.
q If you observe a Pareto effect, focus your improvement efforts
on those few factors.
 - Pareto Analysis
Example:
q A factory team has prepared the following Pareto charts to
address the rising number of customer complaints in a way
management can understand.
Type of Customer Complaints Type of Product Complaints Type of Document Complaints

40 20 10

30 15 7.5

20 10 5

10 5 2.5

Product Docs. Package Delivery Scratch Dent Pin-hole Tear-off HME Other Info. Invoice Wrong Other
missing error quant.

The results suggest that they can solve the majority of the problem
by concentrating on the vital few.
 Scatter Plots/Diagrams
• Scatter plots take place on an X and Y graph
• Which ever variable is on the bottom should
be the dependent variable
• This means that the Y variable changes
according to changes in X.
• Scatter plots are useful for finding direct or
indirect relationships between variables
which can then be used to analyze/improve
quality.
Acme Pizza (Scatter Diagram)
Minutes Cooking Defective Pies
10 1
45 8
30 5
75 20
60 14
20 4
25 6

In this simple example, you can find the existing relationship without
much difficulty but…
 Scatter Diagrams
•Easier to see direct 25

relationship 20

Defective Pizzas
15

10

0
0 20 40 60 80

Time Cooking (minutes)

 Scatter Diagrams
As a quality tool
• What does this tell Acme management about
their processes?
• Improvements?
25

20

Defective Pizzas 15

10

0
0 20 40 60 80

Time Cooking (minutes)