Transcript chapter 1 - GEOCITIES.ws

CHAPTER 7
STATISTICAL PROCESS CONTROL
THE CONCEPT

The application of statistical techniques to determine
whether the output of a process conforms to the
product/service design - to prevent poor quality
 Control charts are used to detect production of
defective products/services OR to indicate that the
production process has changed and that the products
or services will deviate from their design specification
unless something is done to correct the situation
SOURCE OF VARIATION
Nothing can be done to eliminate variation in process
output completely, but management can investigate the
cause of variation to control and minimize it
 common causes : purely random, unidentifiable and
unavoidable, results in symmetric distribution
 assignable causes : variation-causing factors that can be
identified and eliminated, such as unskilled employees
and a needing repair machine which affect the
mean/plan, spread and shape (skewed distribution)
SOURCE OF VARIATION (continued)

A process is said to be in statistical control when
the location (of mean/plan), spread or shape of its
distribution does not change over time
 SPC procedures are used after the process is in
statistical control to detect the onset of assignable
causes so that they can be eliminated
THE INSPECTION PROCESS
Quality measurement
 how to measure quality characteristics (variable,
attribute)
 sample size to collect
 at which stage in the process to conduct inspections
QUALITY CHARACTERISTICS

variables : weight, length, volume that can be
measured
 attributes : can be quickly counted for acceptance,
simple yes-no decision, quality specifications are
complex, measuring by variables are difficult and
costly
 control charts to establish the control limits and to
monitor the process
SAMPLING

Complete inspection : the cost of passing defects to the
next workstation or external customers outweigh the
inspection cost, the need of automated inspection
equipment for accuracy and saving time
 Well-conceived sampling can approach the same degree
of protection as complete inspection, randomly selected,
inspection costs are high
 Zero defects (parts per million) oriented  sampling
plan that attempts to minimize the possibility of wrongly
rejecting good items or wrongly accepting bad items
SAMPLING (continued)

Larger sample sizes are for attribute charts because
more observations are required to develop a usable
quality measure
 Variable control charts require smaller sample sizes
because each sample observation has provided usable
information
 The first and last sample in a small production lot
 Samples come from a homogeneous source (e.g.:
separated samples from different machines, shift, etc)
CONTROL CHARTS

A time-ordered diagram where to plot the quality
characteristics taken from the samples
 A sample statistic that falls between upper and lower
control limit indicates that the process is exhibiting
common causes of variation, while those that fall
outside indicate any assignable causes of variation
 Observations falling outside the control limits do not
always mean poor quality and the ones inside the
control limit may indicate any “warnings”
TYPE I & II ERROR

Control charts are not perfect tools for detecting shift in
the process distribution because they are based on
sampling distribution
 Type I error occurs when the employees conclude that
the process is out of control based on a sample result that
falls outside the control limits, when in fact it as due to
pure randomness
 Type II error occurs when the employee conclude that
the process is in control and only randomness is present,
when actually the process is out of statistical control
TYPE I & II ERROR (continued)

For increasing standard deviation of sampling
distribution, the probability of type I error is lower
while the probability of type II error is higher
 Applying high (3) sigma limits when the cost of
searching for assignable cause is large relative to the
cost of not detecting a shift in the process average
 Using 2 sigma limits when the cost of not detecting a
shift in the process average exceeds the cost of
searching for assignable cause
LOCATIONS OF INSPECTION PROCESS
The considerations :
 cost of inspection to achieve good quality, cost of
poor quality and the their importance
 skill and technology required
 inspection spots (raw material, work in process,
finished product/service)
 customer’s involvement
 the effect on productivity