Operations

Management and Total

Quality Management:
Statistical
Process
Group 7
Control
Jamilla van Gestel
Nathalie Mickael
Monjardin
Jhozel Maye Madelar
Leila Marie Aguilar
08/17/2025Annual Review 1
 Statistical
Process
• Control
Monitori
ng
• Controlli
ng
• Improvin
08/17/2025Annual Review
g 2
Contents
 Control Charts
 Variation in Process Quality
 Uncontrolled Variation
 Controlled Variation
 Control Charts for Attributes The Process Control
 Control Charts for Variables Process Knowledge
 Variable Control Charts for  Acceptance Sampling
Subgroup Data
 Operating
 Variable Control Charts for Characteristics (OC)
Individuals Data Curve

08/17/2025Annual Review 3
Control
Charts
 Used to routinely monitor quality
 Two basic types of control charts:
 Univariate control chart –
graphical chart of one quality
characteristic
 Multivariate control chart –
graphical chart of a sum of
quality characteristics

08/17/2025Annual Review 4
Control
 Contents of the chart:

Charts
 Center line – mean value for the in-
control process
 Two horizontal lines – upper control
limit (UCL) and lower control limit
(LCL)
 Almost all data points fall within
these limits as long as the process is
in control
 Two common uses in an improvement
project:
 Most common universal application –
tool to check process stability and
control
 Less common but more powerful –
use as an analysis tool
08/17/2025Annual Review 5
Elements of a Control
Chart
3 Main Elements
Control limits make sure
1. Stars with a time series time is not wasted looking
graph. for unnecessary trouble.
2. A central line (x) as visual They are computed by:
reference for detecting 1. Estimating the standard
shifts or trends – called deviation of the sample
the process location. data
3. Upper and lower limits 2. Multiplying that number
(UCL and LCL) are by 3
computed from obtainable 3. Adding (3 x to the
data and placed average) for the UCL and
equidistant from the subtracting (3 x from the
central line – position
08/17/2025Annual Review
average) for the LCL 6
dispersion.
Elements of a Control
Chart
Mathematically, it looks like: ^
 ^ – estimate of standard
deviation, calculated from a
sample of data
 – population of standard
deviation, calculated from the
entire population
 Control limits are calculated
from process data, they are
independent of changes in
customer expectations or
varied specification limits.
08/17/2025Annual Review 7
 Control rules take advantage of the
normal curve in which 68.26% of all
data is within plus or minus 1
standard deviation from the average,
95.44% of all data is within plus or
minus 2 standard deviations from the
average, and 99.73% or data will be
within plus or minus 3 standard
deviations from the average. Such
data should be generally distributed
when using control charts, or the
chart may signal unexpectedly high
rate of false alarm.
08/17/2025Annual Review 8
Variations in
Process
Quality
Variation is a normal feature of the
process characteristics where some Data will
are statistically random, while others
are appropriate to the actions of always
production workers.
When workers do not work constantly,
demonstrate
they bring in nonrandom variations
that diminish product quality. The
variation.
input for using variation as a tool in
TQM is to discover random variations,
which are normal to process, an those
due to worker inconsistency. TQM
reduces the nonrandom variations by
acting to reduce worker inconsistency.
08/17/2025Annual Review 9
 Is the variation
typical for the
process or
unanticipated?
How can it be
improved?

08/17/2025Annual Review 10
Produced by Common
Causes
Variation that is  If they are producing excessively much
typical or variation in the system, improvement is
necessary.
customary
 Since they generate the normal everyday
 Driving to work: variation, improvement would entail
• Traffic lights systematic changes.
• Weather conditions
 Driving to work:
• Using different roads
• Departing at a different time
• Changing the mode of transportation
• More extreme – changing residence

08/17/2025Annual Review 11
Produced by Special
Causes
Variation that is strange  If they are producing problems
in the system, they will be specific
or unanticipated events such as an automobile
 Driving to work: breakdown.
• Breakdown of the car  Driving to work:
• Short term action – tow the car to a
• Involvement in an garage
accident • Long term action – get better
 A consequence of a maintenance of the car
particular change and are  The action must happen right away
often connected with a to rise above the special cause and
the long-term action must be taken to
disorganized problem; an avoid recurrence.
accident.  They are not always negative, can be
08/17/2025Annual Review positive too. 12
How do we do it right all
theControlling
Over time? Under Controlling
 Assumption that a special  Assumption that the process
cause is occurring when or system is operating
only normal variation is in normally when something
attendance. special is occurring.
 Learning to drive:  Learning to drive:
• Lane change may be • Takes car out in the rain
traumatic and drives at speed limit
• Hitting the brakes when • Applies brakes at stop signs
a car turns ahead • Skids helplessly into the
intersection
08/17/2025Annual Review 13
 Process without Control
special causes/ Statistical chart
variations of only stability predictabilit
common causes y

 If the system remains the same, data produced will vary normally
between the control limits, and will have the same average as that
shown on the control chart.
 Powerful aspect of control charts – ability to assess process
variation and system performance, facilitates the prediction of
future performance.
 It is generally advisable to eliminate special causes to achieve a
statistically stable process before addressing common causes to
result in Review
08/17/2025Annual the best use of resources and the best improvements. 14
 Uncontrolled
Variation
 Once TQM has identified the
nonrandom variations, it
examines their source by
looking for inconsistencies
that result from uncontrolled
actions.
• Product failure = loose
bolt = worker not
tightening the bolt
consistently
• Product color variation =
 TQM introduces
paint controls to reduce inconsistency and turn uncontrolled
composition
variations
variationinto controlled ones by:
• Training the worker to use a torque wrench
• Specify a single supplier of paint
08/17/2025Annual Review 15
 Controlled
Variation
 When TQM introduces control
to reduce inconsistency, it
must check to see if the
controls reduced the
variations initially observed.
• If use of torque wrench
eliminates product failure,
the controls have
eliminated the variation.
• If product color still varies,
the variation is still
uncontrolled. It may be that the inconsistency is caused by
inconsistent drying time. If it increases color insistency, the variation is
controlled and quality increases with respect to the desired
characteristics.
08/17/2025Annual Review 16
 Control Charts for Attributes
• The Shewhart chart plots quality characteristics that can be
measured and expressed numerically in terms of weight,
height, position, thickness, etc.
• If one cannot represent a particular quality numerically, or if
it is impractical to do so, he often resorts to using a quality
characteristic to sort or classify an item that is inspected in
one of two “buckets”.
• Conforming vs nonconforming units
• Non-defective vs defective units
• Quality characteristics of the said types are called attributes.
• Number of failures in production run, proportion of malfunctioning
wafers in a lot, number of people eating in a cafeteria on a given day,
etc.
08/17/2025Annual Review 17
 C - charts (for count)
• Control charts dealing with the number of defects or
nonconformities
• Used to evaluate process stability when counting defects
or errors and when sample size is constant:
• Injuries per month
• Falls per month

08/17/2025Annual Review 18
 P - charts (for proportion)
 Control charts dealing with the proportion or fraction of defective product
 Used to evaluate process stability when counting defective items and when
sample size varies:
• Total number of circuit boards, meals, or bills delivered varies from one
sampling period to the next
 Each item is only counted once:
• Car doors that need to be repainted, bills that are wrong, incomplete
shipments, and c-sections in a delivery

08/17/2025Annual Review 19
 U - charts (for unit)
• Control charts dealing with the average number of
nonconformities per unit of product
• Used to determine the stability of “counted” data:
• Errors per bill and dents in a car door when there can be more
than one defect per unit and the sample size varies

08/17/2025Annual Review 20
Control Charts
for Variables
 Variables control chart plot continuous
measurement process data, such as
length or pressure, in a time-ordered
sequence.
 In contrast, attribute control charts
plot count data, such as the number of
defects and defective units.
 Variable control charts, like all control
charts, help identify causes of
variation to investigate, so that one
can adjust the process without over-
08/17/2025Annual Review 21
controlling it.
Control Charts for
Variables
Charts for Data Collected in Subgroups Charts for Individual Measurements

 Group of units produced under the  Used when measurements are

same set of conditions. expensive, production volume is low,
• If one wants to chart a particular or products have a long cycle time.
measurement from a process • To test impact strength of parts
• If he collect and measure five parts (destructive testing)
every hour, his subgroup size would • Each point of the graph represents
be 5 individual measurement; thus, the
 Variable control chart for subgroups subgroup size is 1
include:  Individuals control chart include:
• Xbar chart • I chart
• R chart • MR chart
• S chart
08/17/2025Annual Review 22
• Zone chart
 Charts for Data Collected
in Subgroups: R Chart
Plots the process
mean over time.
Used to track the
process level and
detect the presence of
special causes
affecting the mean.

08/17/2025Annual Review 23
 Charts for Data Collected
in Subgroups: Xbar Chart
Plots the process
range over time.
Used to track the
process variation and
detect unexpected
variation.

08/17/2025Annual Review 24
 Charts for Data Collected
in Subgroups: S Chart
Plots the process
standard deviation
over time.
Used to track the
process variation and
detect unexpected
variation.

08/17/2025Annual Review 25
Charts for Data
Collected in Subgroups:
Zone
Plots theChart
cumulative scores based on “zones” at 1, 2, and 3
standard deviations from the center line. Used to detect
unexpected variation.

08/17/2025Annual Review 26
 Charts for Individual
Measurements: I Chart
Plots individual
observations over
time.
Used to track the
process level and
detect presence of
special causes.

08/17/2025Annual Review 27
 Charts for Individual
Measurements: MR Chart
Plots the moving
range over time.
Used to track the
process variation and
detect presence of
special causes.

08/17/2025Annual Review 28
 The Process Control
• Is a mixture between the statistics and engineering
discipline that deals with the mechanism, architecture,
and algorithms for controlling process. Some examples
are:
1. Controlling the temperature of a water stream by
controlling the amount of steam added to the shell of
a heat exchanger
2. Operating a jacketed reactor isothermally by
controlling the mixture of cold water and steam that
flows through the jacket of a jacketed reactor
3. Maintaining a set ratio of reactants to be added to a
reactor by controlling their flow rates
4. Controlling the height of fluid in a tank to ensure that
it does not overflow
08/17/2025Annual Review 29
 The Process Control
• Generally, anything that requires continuous monitoring of
an operation involves the role of a process engineer and
are helped by the Distributed Control System (DCS) that
communicates with the instruments in the field since
today’s chemical plants have gone to full automation.
• Safety of workers and the community around a plant
should be, if not already, the number one concern for
most engineers as they begin to design their process.
• Chemical plants have great potential to do severe damage
if something does wrong and it is inherent the setup pf
process control to set boundaries on specific unit so they
don’t injure or kill workers or individuals in the community.
08/17/2025Annual Review 30
 The Process Control
A control system is required to perform either or both
tasks:
1. Maintain the process at the operational conditions
and set points
2. Transition the process from one operational
condition to another. Reasons why a process may be
moved from one operational set to another:
a. Economics
b. Product
specifications e. Consumer/ customer
c. Operational specifications
constraints f. Environmental
d. Environmental
08/17/2025Annual Review regulations 31
Process Knowledge
 In recent years the role of process control and SPC in
particular, has changed increasingly and is seen not just
as a convenient method of keeping processes in control,
but also as an activity which is fundamental to the
acquisition of competitive advantage.
 Traditionally, it was seen as one of the most
operational, immediate and “hands-on” operations
management techniques. Yet it is not being connected
with an operation’s strategic capabilities.

08/17/2025Annual Review 32
Process Knowledge
 Logic of the argument:
• SPC is based on the idea that process variability indicates whether
a process is in control or not
• Processes are brought into control and improve by reducing
process variability. Eliminating the assignable causes of variation
• One cannot eliminate assignable causes of variation without
gaining better understanding of how the process operates.
Involves learning the process on an increasingly detailed level
• Learning = process knowledge is enhanced = operations
managers are able to predict the process performance under
difficult circumstances and has greater capability to carry out
tasks at higher level performance
• Increase process capability is difficult for competitors to copy. It
cannot be bought “off-the-shelf” and only comes from time and
effort being invested in controlling operations processes leading to
08/17/2025Annual Review 33
strategic advantage.
Process Knowledge
Process
knowledge
enhancement
Process
and builds
control
uncopiable
process
capability
08/17/2025Annual Review 34
 Acceptance Sampling
 Popularized by Dodge and Romig and originally applied by the U.S military to test bullets
during WWII
• If every bullet was tested in advance, no bullets would be left on the ship
• If none were tested, malfunctions might occur in the field of battle, with potentially
disastrous events
 Dodge:
• A sample should be picked at random from the lot and a decision should be made
regarding the disposition of the lot
• Lot Acceptance Sampling or Acceptance Sampling → the decision is either to accept or
reject the lot
 Acceptance sampling is a major component of quality control and is useful when the cost of
testing is high compared to the cost of passing a defective item or when testing is
destructive
• A compromise between doing 100% inspection and no inspection
• May be applied where large quantities of similar items or large batches or material are
being bought or are being transferred from one part of an organization to another
• “Middle of the road” approach between no inspection and 100% inspection.
• Two major classifications or acceptance plans:
 By attributes
08/17/2025Annual Review (“go, no-go”) → most common 35

 Its main purpose is to decide
whether or not the lot is
acceptable, not to estimate the
quality.
 It is employed when one
or several of the
following hold:
• Testing is destructive
• The cost of 100%
inspection is very
high
• 100% inspection
takes too long
08/17/2025Annual Review 36
Operating
Characteristic (OC)
Curve
Depicts the discriminatory
power of an acceptance
sampling plan.
 Plots probabilities of
accepting a lot versus the
fraction defective.
 When OC curve is plotted,
sampling risks are obvious.
• Should always be examined before
using a sampling plan

08/17/2025Annual Review 37