Chapter 9 Flashcards
The Seven Tools of Quality Control
Flowchart
Cause and Effect
Check Sheet
Scatter Diagram
Histogram
Control Chart
Pareto Chart
Flow Charts
Understanding the process and identifying possible problem areas
Check Sheets
Tabulating data on the problem area
Histograms
Illustrating the frequency of occurrence of measures
Pareto Diagrams
Identifying the most important problems
Cause and Effect Diagrams
Showing possible causes of the problem
Scatter Diagram
Investigating causes and effects
Control Charts
Holding the gains from process improvement
Specification limits
deal with individual product characteristics
Control limits
deal with sample process characteristics (i.e., means and ranges)
Control Limits and Specification Limits
Should never be plotted on the same graph
Process Quality Control
Assignable causes -Can be identified and corrected Common (random) causes -Occur randomly -Cannot be changed unless process is redesigned
Basic assumptions of Process Quality Control:
Every process has random variation in it.
Production processes are not usually found in a state of control (they have assignable causes of variation present).
“State of Control”; what does it mean?
Sources of unnecessary variation (assignable causes) have been eliminated
Remaining variation is due to random causes
Steps in Designing a Statistical Process Control System
- Identify critical points to apply control (where data should be collected)
- Identify the critical quality characteristics of that aspect of the process & Decide on the type of measurement (what)
- Decide on the amount of data to be collected (how much)
- Decide who should collect the data, construct the charts, and apply the charts (who)
Identify critical points to apply control (where data should be collected)
Processes impacting incoming materials & services
Internal processes (especially prior to high cost processes and processes that are “sensitive”)
Identify the critical quality characteristics of that aspect of the process
& Decide on the type of measurement (what)
Variable data
Attribute data
Variable data
Process (or product/service) characteristic that can be measured on a continuous scale:
-Length, size, weight, height, time, velocity, temperature, etc.
Typically mean & range or standard deviation
Attribute data
Process (or product /service) characteristic evaluated with a discrete choice:
-Good/bad, yes/no, 0/1, ratings, counts, proportions
Decide on the amount of data to be collected (how much)
Sample size
- Large enough to detect deviations
- Variable data allow for a smaller sample size
How often to sample?
-Depends upon inherent variability of the process, cost, production rate
x-bar Chart
Used to control the mean value of a particular characteristic of a process
R Chart
Used to control the variability of that same characteristic of the same process for which the x-bar chart is being used to control the mean
SPC - Variables - Example
Characteristic: Critical dimension of a process
Before collecting the data, make sure the process is “in control”!
Action Required for Out-of-Control Indicators - Variables
- Stop the process
- Conduct an investigation to determine the assignable cause*
- Take corrective action to remove the assignable cause
- Resume the process
* If no assignable cause is found, assume a Type I error has occurred & jump to Step #4
Type I error
(a.k.a. producer’s risk and a error) - The control chart for a process indicates the process is “out-of-control” when the process is actually “in-control” and, thus, the process is stopped unnecessarily
Type II error
(a.k.a. consumer’s risk and b error) - The control chart for a process indicates the process is “in control” when the process is actually “out-of-control” and, thus, the process is allowed to continue when it should be stopped.
Action Required for Out-of-Control Indicators - Attributes
1.Stop the process1
2.Conduct an investigation to determine the assignable cause
3.Take corrective action to remove the assignable cause2,3
4.Resume the process
1If sample value falls below the LCL, only if needed to conduct the investigation for assignable cause
2If no assignable cause is found in Step #2, assume a Type I error has occurred & skip this step
3If downward shift is indicated, attempt to ingrain assignable cause into the process
Acceptance Sampling - General
A random sample is drawn from a lot (population)
The sample is inspected to determine the number of defective items in it
The number of defective items found is compared to a cut-off value
The result from the (representative) sample is applied to the entire lot
Quality Control in Industry
75% use process control charts
More use of variable (x-bar and R) charts than attribute (p and c) charts
“The Seven Tools of Quality Control”
Quality control in the service industry
The Seven “M’s” of a Process
Material Manpower Methods Machines Measurement Maintenance Management
