You are on page 1of 29

Statistical Process Control

Statistical Quality Control

SQC is an optimization philosophy concerned with


continuous process improvements, using a
collection of (statistical) tools for data and process
analysis and making inferences about process
behavior and decision making
Statistical Quality Control
• Statistical quality control can be divided into three broad categories:

1. Descriptive statistics are used to describe quality characteristics and


relationships. Included are statistics such as the mean, standard deviation, the
range, and a measure of the distribution of data.

2. Statistical process control (SPC) involves inspecting a random sample of


the output from a process and deciding whether the process is producing
products with characteristics that fall within a predetermined range. SPC
answers the question of whether the process is functioning properly or not.

3. Acceptance sampling is the process of randomly inspecting a sample of


goods and deciding whether to accept the entire lot based on the results.
Acceptance sampling determines whether a batch of goods should be accepted
or rejected.
How is SQC implemented
• Descriptive statistics are used to describe certain quality
characteristics, such as the central tendency and variability
of observed data.
• Although descriptions of certain characteristics are
helpful, they are not enough to help us evaluate whether
there is a problem with quality.
• Acceptance sampling helps us decide whether desirable
quality has been achieved for a batch of products, and
whether to accept or reject the items produced.
How is SQC implemented
• To help us identify and catch a quality problem during the
production process, we need tools in the statistical process
control (SPC) category
Why we need process control
• No two products are exactly alike because of
slight differences in materials, workers,
machines, tools, and other factors.
• These are called common, or random, causes
of variation.
• Common causes of variation are based on
random causes that we cannot identify.
• These types of variation are unavoidable and
are due to slight differences in processing.
Causes of variations
• Variations due to common causes
– have small effect on the process
– are inherent to the process because of:
• the nature of the system
• the way the system is managed
• the way the process is organized and operated
– can only be removed by
• making modifications to the process
• changing the process
Causes of variations
• Variations where the causes can be precisely identified and
eliminated.
• These are called assignable or special causes of variation.
• Variations due to special causes are
– localized in nature
– exceptions to the system
– considered abnormalities
– often specific to a
• certain operator
• certain machine
• certain batch of material, etc.
Investigation and removal of variations due to special causes
are key to process improvement
Using SPC tools for SQC
• Statistical process control methods extend
the use of descriptive statistics to monitor
the quality of the product and process.
• Normal distributions with varying Differences between symmetric and
standard deviations skewed distributions
Using SPC tools for SQC
• Using statistical process control - determine the amount
of variation that is common or normal.
• Then monitor the production process to make sure
production stays within this normal range.
• Ensure the process is in a state of control/ is stable.
• The most commonly used tool for monitoring the
production process is a control chart - Control charts are
used to detect whether a process is statistically stable
Control Charts
• Are an important aid or statistical device used for the
study and control of repetitive processes - Developed by
W.A.Shewhart.

• Control limits on the charts are so placed as to disclose


the presence or absence of the assignable causes of
quality variations.

• The objective of the quality control system is to minimize


the unassignable causes of variations and eliminate the
assignable causes of variation.
Control charts are run charts with superimposed normal
distributions. Control charts provide a graphical means for

testing hypotheses about the data being monitored.


Control Charts
• The procedure behind the application of control charts is
– sample the process at regular intervals
– plot the statistic (or some measure of performance), e.g.
• mean
• range
• variable
• number of defects, etc.
– check (graphically) if the process is under statistical
control
– if the process is not under statistical control, do
something about it

Control charts: types of charts


Types of data
• Statistical data can be characterized as
either attribute data or variable data.
• Attribute or discrete or countable data-data
on a characteristic that can assume certain
distinct values. It records the no. of articles
that are conforming or non-conforming to a
specified requirement
• Variable or continuous data is one when a
record is made of the actual measured
quality characteristic such as a dimension.
Different control charts are used depending on the nature of the data
• Commonly used charts are:
· for continuous (variables) data
o Shewhart sample mean( - chart)
o Shewhart sample range (R-chart)
o Cumulative sum (CUSUM)
o Exponentially Weighted Moving Average (EWMA) chart
o Moving-average and range charts
· for discrete (attributes and countable) data
o sample proportion defective (p-chart)
o sample number of defectives (np-chart)
o sample number of defects (c-chart)
o sample number of defects per unit (u-chart)
exercise

x-bar chart
Procedure to construct a control
chart

K = number of samples
Measurements collected from LINE 1 for Day 1- at critical point -Chest(Size M running)

time 10 a.m 12 noon 2 p.m 4 p.m 6 p.m

Bundle 1-Op1 1 40 41 40.75 41 40.75


2 41 41 41 40 41.25 Calculate
3 41.25 40.75 40.5 40.75 40.5
4 40.75 41 41.25 40.5 41 the UCL and
5 40.5 41 41 40.5 40.5
40.7 40.95 40.9 40.55 40.8
LCL and
Make the
Bundle 2-Op1 1 41.25 40.75 41.25 40.75 41 control chart
2 41 40.75 41.25 41 41.25
3 41 40.5 40.75 40.75 41.25 On excel
4 40.5 40.75 41 41 40.75
5 40.75 40.5 41.25 40.5 41 sheet.
40.9 40.65 41.1 40.8 41.05

Bundle1-Op2 1 41.25 41.25 40.5 41 41


2 40.5 40.75 40.75 40.75 41.25
3 40.75 41 41.25 40.75 41
4 40.5 40.75 40.75 41 41.25
5 41.25 41.25 40.75 40.75 40.75
40.85 41 40.8 40.85 41.05

Bundle2 Op2 1 41 40.5 40.75 40.5 41.25


2 40.75 41 41 41.25 41
3 41 41 41.25 41 41.25
4 41.25 41 40.5 40.75 41
5 41 40.5 41 40.75 40.75
41 40.8 40.9 40.85 41.05
SPC implementation in the apparel industry
• there are 2 types of charts being used
• wormanship control chart or attribute chart
• measurement control chart or variables chart

• WORKMANSHIP OR ATTRIBUTE CHART


PROCEDURE
• these charts are used for visual inspection.
• these charts can be used for fabric, trim, cutting, in-line & final
inspection areas.
• the usage of this chart is determined during the pre-production
meeting.
SPC
• workmanship or attribute chart procedure-

• during the pre-production meeting, the QA mgr & his/her team will decide as to
where the workmanship or attribute charts will be positioned based on the
critical operations of the garment.

• the locations of measurements or variable charts are also determined at this


meeting.

• critical operations are locations of a garment where the quality team predicts
having potential problems due to difficulty in construction etc.

• the auditor must have acceptable standards set by the mgr in order to conduct
an audit.

• the sealed sample can be used as a standard.


SPC
• if necessary a mock-up of the operation (part of a garment) can be
established at the crtical operation point.

• the auditor must audit on an hourly basis.

• the auditor could use aql 2.5 single sampling plan in order to select sample
size based on hourly production output.

• “n” is identified as the sample size. “np” is identified as the number of defects
found. “p” is the proportion.
“np”
____ = “p” p x 100= % of defects.
“n”
SPC
• the upper control limit (ucl) for this chart is determined by
the target rejection rate planned for a particular period (
e.g. 5%).
• the ucl is a fixed control limit for in-line audits.
management can decide if they want to lower the defect %
by bringing the ucl down.
• for each out of control point, a corrective action plan needs
to be recorded by answering what happened, how did it
happen, how was it corrected & who corrected?
SPC
• measurement or variable chart procedure:-

• 5-7 garments are measured per size & per hour at each critical point.

• upon measuring, the difference is recorded & plotted onto the chart.

• in general 1 / 2 to 3/4 of customers tolerance range is recommended for use


during in-line audits.

• e.g. – since in general, we want waist measurement in a 5


pocket jeans to be more on the plus side than on the minus side, we can have
the ucl at + 1 / 2” & the lcl at – 1 / 4”. the range in this case is 3 /4”.

• if 1 out of 7 measurements reaches 1/8” out of the control line, an additional


7 measurements are taken & the difference plotted below the hour column
without joining the dots. the appearance would be more like a scattered
chart.
SPC
• if none of the 7 measurements are out of the control line, then
the auditor must return in the next hour. if one or more
measurements are out of control line, then corrective action
must be taken which could include 100% measurement of
the bundle or size.

• if 2 out of 7 measurement reaches 1 / 8” out of control line or if


1 or more measurements reaches 1 / 4” out of control line, then
20 additional garments must be measured & plotted in a similar
fashion as stated in the above point.
SPC
• if one or more measurements from the 2o is out of
tolerance by 1 / 4” or more, then corrective action must
be taken.

• in this chart we can also determine the range of


each size as well as for each day.
Spc new format – excel recording

You might also like