STATISTICAL PROCESS
CONTROL
VARIABLE AND ATTRIBUTE CHARTS
Presented By :
Surya Bakshi
Statistical Process Control is
a technique used within the TQM framework
Aims to reduce variation in processes
control, manage, analyze and improve the
performance
eliminates special causes of variation
defective products,
operator error,
errors in measurements
CONCEPTS
PROCESS
Is a set of interrelated or interacting
activities which transform inputs into
outputs.
VARIATION
No two products are exactly alike due to
sources of variability
Sources of variation : speed of the
operator, accuraccy, procedure followed,
variation over time.
Advantages of maintaining
a state of Statistical Control
Variation( inherent) is restricted to
common causes
Since variability exhibits a regularity in
pattern, the process is repeatable
Since process is repeatable, quality of
future production can be predicted.
SPC is a strategy of Defect prevention which aims
To increase the knowledge about any process
To steer a process to behave in the desired way
To reduce variation in the product or in other ways to
improve performance of the process
To reduce cost and increase profit by improving the
performance of the process
Increase consistency of the process
Periodic sampling of products from the
process and inspecting them for
characteristic under cosideration
Computing a suitable index of quality for
each statistic
Taking actions to detect and eliminate
special causes wherever control limits are
exceeded
Leaving the process alone when there’s no
evidence of “out of control”
Control charts or graphical trend
analysis are used to understand
changes in the process mean and
process standard deviation.
Process variation
adds needless cost to the product
reason why a product does not comply
with specifications
are a graphical device to aid in process control
Control chart consists of:
Points representing a quality characteristic in samples under
observation
A Centre line, drawn at the process characteristic mean,
which is calculated from the data.
Upper and Lower Control Limits that indicate the threshold at
which the process output is considered statistically 'unlikely'.
Control Chart technique can be used for
both Variable and Attributes.
Measurement or Variables
quality characteristics which can be measured
Attribute
quality characteristics that can be evaluated
with a discrete response
not measurable on the quantitative scale
Data obtained by classification or counting
are called
There are four types of attribute charts used for
non measurable defect controlling and
defectives controlling
N and NP Charts for the control of number of
defectives or non- conforming
C and U – Charts for the control of the defects or
non- conformities
• Used at places where
▫ various defects could occur in a single product that is being
manufactured
• Fixed sample size , 13 in this case as per 2.5 AQL
• The proportion of defectives are
▫ plotted and
▫ analyzed for improvement of the process
The graph is made from the quality checks done in the
front section of Jackets Line 3 for three consecutive days
The plotting shows that the process is shown to be
within limits and
very much stabilized
the UCL is far beyond the fraction of defectives that have
been plotted.
Hence if this chart is applied it will
not solve the purpose
identifying the process instability is difficult.
Used at places where
various defects could occur
product wise checking
variable sample size
The UCL and LCL are calculated
on hourly basis
on the last hour’s data available
The chart is
very time taking to be prepared
The Quality Checker has to be trained
Follow up to let them know how to prepare it.
The same problem of data always show within
control persists
Hence, the chart cannot be used.
The two chart applicable to check no. of
defectives is ruled out.
The C Chart and the U chart are used to assess defects
occurring in the product individually.
The U chart is used at places where
The defects are checked to be present as per a specific unit,
For example is carpet making
The defect per square unit area is checked in this case.
There is no such unit that can segregate the garment defects
hence U chart is also not applied in this case.
• The C chart loses its characteristic of showing
▫ the no. of defects present
▫ the fractional values are plotted
• The C Chart previously
▫ Was used for several defects
▫ It is difficult to asses as to which defect led to the
inconsistency
▫ Hence should be used for a single defect.
Preparation of separate chart for different defect that has to be
checked
The no. of defects instead of fractions should be plotted on
hourly basis
the UCL and LCL for the next day to be calculated on the basis of
the last day’s data
THE NEW CHART proposed to be implemented.
The new chart
shows the situations out of control where
the defect needs to be checked
makes it easier to identify the root cause of
the defect
when it happened
where it happened
who did it and
proper steps can be taken to mend it, and
record is maintained for the future
reference
