Control Charts are tools for tracking variation based on the principles of probability and statistics

1
SPC Statistical Process Control

• Control Charts are tools for tracking variation
  based on the principles of probability and
  statistics

2
Variation

• Exists in any process
• error rate made by receptionist entering guest
  record data,
• bus time between two points,
• ounces of beverage in a bottle,
• number of minutes past the alarm that you stay in
  bed in the morning.

3
Sources of Variation 2 Types

• Common (random) causes chance or generally
  unidentifiable sources of variation.
• Slight variation in walking speed
• Slight variation in raw material
• Controllable or Assignable causes A reason why
  the change occurred
• people blunder, faulty setup, or a batch of
  defective raw material,worn equipment,
  fluctuating temperature
• Super huge peanuts arrived from supplier so we
  only got 8 in each bag.

4
Controllable Causes
Average
Grams
(a) Location
5
Controllable Causes
Average
Grams
(b) Spread
6
Controllable Causes
Average
Grams
(c) Shape
7
The Normal Distribution
? Standard deviation
8
How variation impacts our process

• Generally random variation cannot economically be
  eliminated from a process.
• Controllable variation can be detected and
  elimination of its causes is economically
  justified.
• Observations beyond the control limits are
  attributed to controllable variation.

9
Process Control Chart Activities

• Periodically sample from our operation or
  process.
• Calculate some characteristic like average,
  standard deviation, or range.
• Plot the characteristic in time order on the
  chart.

10
Purpose of Charts

• To ensure the process variation is in control
• To ensure that the process is capable of meeting
  the requirements (specifications and tolerances
  of the organization)

11
Variable Control Charts (X bar R)

• Measurement charts some characteristic we can
  measure (weight, time, distance)
• X average measurement for the sample
• R range of the measurements in the sample
• Variable charts have lots of information, better
  for advanced analysis of a process

12
Control Limit Formulas Constants (A2, D3, D4)
In Textbook pg. 167
Our sample size n ?
13
Control Chart Examples
UCL
Nominal
Variations
LCL
Sample number
Appears to have normal variation
14
Control Chart Examples
UCL
Nominal
Variations
LCL
Sample number
A process with a gradual trend
15
Control Chart Examples
UCL
Nominal
Variations
LCL
Sample number
Points Outside of the Control limit
16
Attribute Chart P-Charts

• Number of defective units in a sample. Yes/No,
  Pass/Fail, Go/No Go criteria
• p- easy to measure (pass/fail) but sample size
  must be big enough to detect at least one
  defective item on average
• ppercentage faulty in sample
• N size of the sample

17
Control Charts for Attributes
P-Chart
18
Six Sigma Quality
When a process operates with ?6s variation inside
the tolerance limits, only 2 parts out of a
million will be unacceptable.