TOPIC 7 Flashcards
Statistical Quality Control
uses statistical techniques and sampling to monitor and test the quality of goods and services.
– Acceptance sampling relies on inspection, determines to accept or reject a product.
– Statistical process control determines if process is operating within acceptable limits during production.
Inspection
is an appraisal activity that compares the quality of a good or service to a standard.
Phases of Quality Control
-acceptance sampling
- statistical process control
- continuous improvement and six sigma
When is the amount of inspection optimal
when the sum of the costs of inspection and passing defectives is minimized.
Descriptive statistics
describe the quality characteristics; this
includes Mean, SD, range and measure of distribution of data
Acceptance sampling
randomly inspecting a sample of goods
and deciding whether to accept the entire lot based on the results.
Statistical process control
inspecting a random sample to decide whether the product characteristics are within the certain acceptable bounds (process capability and control charts)
SQC - planning process
- Define important quality characteristics (customer requirements)
- For each quality characteristic,
- Determine quality control point (Critical control points – Inputs; Process; Outputs)
- Plan inspection (Appraisal Cost):
• How to inspect (automatic, manual, …)
• How much to inspect (sample size and sample frequency) – the cost of inspection vs. the cost of passing defects
• Where (on-site or centralized)
Plan corrective action (if process has quality problems then fix it - PDCA)
Process capability
Is the process capable of meeting the user (manufacturer defined)
tolerances, i.e. Upper and Lower Specification Limits?
Process control ( through control charts)
Is the process output stable?
Is the process output within the statistically defined Control Limits?
Process capability
The ability of a process to meet the design specifications Design spec. is usually a range [a, b]
Random (common) variation
caused by countless minor factors.
– Unavoidable
Assignable variation
causes can be precisely identified and eliminated.
Control charts
Control limits are generally set at 3 standard deviations of the process mean.
Abnormal variation most likely furor assignable causes
^^^
Normal variation due to chance
Abnormal variation most likely due to assignable causes
Sampling distribution and the variation limits
For a set of data with a normal distribution, a random measurement will fall within:
± 1 SD 68.3% of the time ± 2 SD 95.5% of the time ± 3 SD 99.7% of the time
Variables
to monitor characteristics that can be measured and
have a continuum of values, such as height, weight, or volume –
Examples: Soft drink bottled quantity; the weight of a bag of sugar; the
temperature of a baking oven; the diameter of plastic tubing
Attributes
to monitor characteristics that have discrete values
and can be counted
– Single decisions (yes/no, acceptable/unacceptable) examples: the apple is good
or rotten, the shoes have a defect or do not have a defect
– Counting examples: the number of broken cookies in a box; the number of dents on the car
X bar and range
X-bar charts measure shift in the central tendency of the process
R (Range) charts monitor the dispersion or variability of the process
What is a good shift in the mean
When the mean in changing and shifting upwards we would consider this a good shift in the mean
Different charts and what the show with the mean and variation
Shifting mean:
X chart: reveal shift
R-chart: does not reveal shift
Increasing deviation
X chart: does not reveal shoft
R chart: reveals shift
P-Charts
to monitor the proportion of defective in a process
– Observations can be placed into two categories:
• Good or bad
• Pass or fail
– Data consists of multiple samples of several observations
– Sample proportion of defective are recorded
Non-random patterns in control charts
Trend: sustained upward or downward movement
Cycles: wave pattern
Bias: too many observations on one side of the centre line
Level shift: a shift in the level
Too much dispersion: the values are too spread out
Trend: sustained upward or downward movement
Cycles: wave pattern
Bias: too many observations on one side of the centre line
Level shift: a shift in the level
Too much dispersion: the values are too spread out
