QUALITY MANAGEMENT

TOOLS III
THE SEVEN BASIC QUALITY CONTROL TOOLS
Basic and effective quality control tools

There are several time-tested established methods for quality control (and for quality improvement) such as:
1. Cause-and-effect (fishbone) diagrams
Identify potential causes for quality problems
2. Flowcharts
Provide a clear picture of how operations or processes work
3. Check Sheets (data collection forms)
Count and classify defects by type or location
4. Pareto charts
Separate the vital few from the trivial many – direct effort to the vital few
5. Histograms
Show how a variable is distributed – e.g., piece size
6. Scatter Diagram
Identify potential correlations between variables
7. Graphs (line graph, bar graph, run charts)
Analyze and visualize the evolution of a variable – control charts are built from run charts
CAUSE AND EFFECT DIAGRAM
Ishikawa diagrams help us brainstorm to identify defect sources
FLOWCHART
Flocharts provide clarity and visibility over a process

The abstraction flowcharts provide enables managers to design ideal processes

• Flowcharts can take many forms

• They are the universal process description language
• Make process simplifications evident
• Show where improvements are possible
• Show error-prone steps needing poka-yoke

Hungry Eat pizza slice Satisfied? End

y
n
CHECK SHEETS
Check sheets are a data collection tool which enable statistical process control

Tally check sheets can be use to track fault types and sources
CHECK SHEETS
Check sheets are a data collection tool which enable statistical process control

The defect-location check sheet enables defects to be located

• The sheet is usually a sketch of the product

or part being manufactured
• The fault location can reveal the defect-
causing step
• Different types of fault can be registered
using different markings
PARETO CHARTS
With the pareto rule we can direct our efforts to high-impact problems

70
(64)
60

50
• Rule inspired in Italy’s wealth distribution
Percent from each cause

40 • 20% of causes lead to 80% of effects

30
• Main use is prioritization
• Very intuitive visual display
20
(13) • Prevents shifting the problem to another step
(10)
10 (6)
(3) (2) (2)
0
 HISTOGRAMS
Histograms show how a variable is distributed

40
35 Histograms help problem solvers recognize three
30 distributions traits:
Frequency of calls

• Center: mean or median should gall at the center of

25
the distribution. Gaps between the two indicate bias.
20
• Width: the range is shown in the histogram. Low
15
width indicates high process predictability.
10
• Shape: the overall shape can indicate problems in
5 the data or influences on the overall distribution.
0
Telephone call duration, min
HISTOGRAMS
Histograms show how a variable is distributed
HISTOGRAMS
Histograms show how a variable is distributed
SCATTER DIAGRAMS
Scatter diagrams show how two variables are related

• The tightness of points plotted on scatter diagrams

give an indication the strength of the relationship
between two variables.
• A cluster or stripe of indicates a strong correlation
between the variables.
GRAPHS
Run charts show a variable’s values throghout time

• They are used to show variation and trends in a

process over time
• They can show process performance declines and
improvements over time.
• They can be used to examine both variables and
attribute data.
• Time is displayed on the x axis and the variable or
attribute being investigated is recorded on the y axis.
• There are no control limits in a run chart.
PROCESS CONTROL
Organizations need to collect data to control quality throghout production stages

Inputs Conversion Outputs

Raw Materials,
Production Products and
Parts, and
Processes Services
Supplies

Control Charts Control Charts

and Control Charts and
Acceptance Tests Acceptance Tests

Quality of Quality of Quality of

Inputs Partially Completed Outputs
Products
 PROCESS CAPABILITY
Ultimately, the process needs to be capable of fulfilling the specifications
PROCESS

Process c
Natural

PROCESS
Natural
Design control
control
specs limits
limits

Process cannot meet specifications Process can meet specifications

PROCESS

Natural Design
control specs
limits

Process capability exceeds specifications

 PROCESS CONTROL
Data should be collected strategically since the collection may be costly

When/where to inspect/control During the Production Process Sampling

Variables (continuous)
➢ Inspect before costly operations.
➢ Inspect before operations that are likely to produce faulty items.
➢ Inspect before operations that cover up defects.
➢ Inspect before assembly operations that cannot be undone. Attributes (binary)

➢ On automatic machines, inspect first and last pieces of production runs,

but few in-between pieces.
➢ Inspect finished products.
PROCESS CONTROL
Data should be collected strategically since the collection may be costly
PROCESS CONTROL
Control charts are used to detect special causes of variation

We want to take corrective action as soon as possible To detect special causes

Common • One datapoint corresponds to a sample

Causes mean and spread
• Patterns in the data show if the source
of variation is common or special
• We want to detect variable\attribute
distribution changes as fast as possible
• We try to use extraneous factor to
improve our process if variability lowers
• We seek corrective action if the sample
mean deviates from target.
Special
Causes
PROCESS CONTROL
We control variables’ position and spread and attributes’ position

Types of measurements:
• variable data (product characteristic that can be measured)
−length, size, weight, height, time, speed, output, etc.
−discrete values
−X&MR (no sampling), Xbar and R (small sample), or Xbar and s charts (large sample)
• attribute data (product characteristic evaluated with a discrete choice)
−good vs bad, pass vs fail, etc.
−binary values
−p or np charts for defectives; c or u charts for defects per unit
Basic assumption -- sample means are normally distributed
PROCESS CONTROL
Control chart patterns show us if the process is out of control
In control Cycle (day and night shifts)

Shift in process average Patterns

PROCESS CONTROL
Control charts are used to detect special causes of variation
CLASS EXERCISE - TAGUCHI’S NAPOLITAN PIZZA
Lets try and implement quality control at the Taguchi’s napolitan pizzeria

Imagine you are using your quality management skills

to open an new pizza place in you hometown. You
have homely place near the city’s historical center. You
have the kitchen assembled and bellboys hired. But
now you need a cook. You set the ad and receive three
CVs.
You make individual 15cm radius pizzas for 10€ each,
around 120 per day. They cost 7€ in materials to
make, if they are of the right size. So, size increases
make it more expensive because of material costs,
and decreases make it equally more expensive
because of customer dissatisfaction.

(1) select the Pizzaiolo, (2) compute EoM profit and (3)
setup control charts for pizza size and quality.