Process control
and
process capability
Performance variability
All products and process performance
measures display variability. Externally
and internally.
• External- this measurement such as
customer satisfaction indices, relative
product ranking etc.varies from market
to market.
• Internal- internally no two on assembly line
are identical.even under identical
circumstances, the time and cost required
to produce and deliver the cost may not be
identical.
Another example can be in a bank customers
conducting two identical transactions may
require different processing time.
Sources for all such variability may be
either internal or external.
Internal can be untrained workman,
imprecise instrument like that while as
external reason can be poor quality of
material, delivery delays etc.
• It can be said that ,variability to a
discrepancy in the actual and expected
performance.
• The process that display performance
variability are generally judged less
satisfactory than those with consistent
and predictable performance.
• Customer satisfaction itself can be
defined in terms of variability between
their expectations and experience of the
product. It may be due to the gap
because of following reasons -
1. What the customer wants & what the
product is designed to do.
2. What the process can produce and what
actually it producing.
3. How is the produced product is expected
to perform and how it is performing.
4. How the produced product actually
performing and how actually customer
perceives it.
Analysis of variability
• Simple graphical methods for collecting,
organizing and displaying information
about the performance of variability.
1. Check sheets
2. Pareto analysis.
3. Histograms.
4. Run chart.
5. Multi-vary charts
1. Check sheets – it is a tally of documents
with flow units documented by type of
defects.
Type of defect Total No. of defects.
a 5
b 3
c 4
d 6
e 2
• Pareto charts – all defects are not
created equal, either in terms of
importance or frequency of occurrence.
so in given limited time managers would
like to identify the critical problems on
which to focus. It is based on the
principle of 80~20.
The 80~20 Pareto principle states that
20% problems amount for 80% of total
defects.
Histogram-
It is a bar plot displaying the frequency
distribution of an observed performance
characteristic.
It summarizes the distribution in terms of
its mean which is defined as the
expected value around which the
distribution is centered.
And the Standard Deviation which
measures the spread of it the
distribution around the mean.
Run charts – a run chart is a plot of
performance measure monitored over time.
It displays variability in process output
across the time.
It helps in identifying the structural variability
such as the trend or seasonality.
X and R charts helps in this area.
Average is measure of expected performance
while as Range R is the difference between
highest value minus lowest value.