Data processing: measuring – calibrating – or testing – Measurement system in a specific environment – Quality evaluation
Reexamination Certificate
2000-05-22
2004-03-23
Bui, Bryan (Department: 2863)
Data processing: measuring, calibrating, or testing
Measurement system in a specific environment
Quality evaluation
C702S081000, C702S082000, C702S182000
Reexamination Certificate
active
06711514
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to methods, apparatus and products for quality control. In another aspect, the present invention relates to methods, apparatus and products for analyzing a quality control regimen. In even another aspect, the present invention relates to methods, apparatus and products for analyzing a quality control regimen comprising one or more quality control tests and determining which tests can be eliminated from the regimen or sampled at a different rate. In still another aspect, the present invention relates to methods, apparatus, and products for analyzing a quality control regimen comprising one or more quality control tests, determining which tests can be eliminated from the regimen or sampled differently, and then conducting “what if” scenarios on the data absent the eliminated tests or at the different sampling rate.
2. Description of the Related Art
Statistical sampling is generally employed where there is a concern with quality control. As a non-limiting common example, electronic components or goods are generally subject to a number of performance tests which are selected and designed to provide an indication of the quality of the components or goods. As it is generally impractical to test each and every component or good, resort is made to statistical sampling.
Statistical sampling generally involves taking small samples from a larger sized grouping, and making a generalization regarding the quality of the entire grouping based on the small sample. The number of samples taken, the size of the samples taken, and the type and number of quality control tests conducted on the samples are all a function of the desired level of confidence required for the quality control process. A cost benefit analysis is generally also applied to the quality control process, that is, the cost of the quality control process must not outweigh the benefits achieved.
For example, for a lot of 5,000 electronic components, a sample of 10 components might be selected on which to run 7 quality control tests. The decision for the entire lot of 5000 components, whether it be “passing” “failing”, “conditional acceptance”, “further testing”, or some other choice, will depend upon the outcome of the seven tests on those 10 sample components.
In general, for any given test(s), the level of confidence regarding the quality can be increased or decreased by respectively increasing or decreasing the sample size selected from the lot. Depending upon the circumstances, anywhere from a small fraction, to a larger fraction to 100 percent of the lot may be tested.
Additionally, while the level of confidence may also be effected by conducting more or fewer quality control tests, the effect on the level of confidence is not so easily determined, as each test contributes differently to the confidence level, and determining that contribution is difficult.
Ideally, after a quality control testing regimen is implemented, inquiry should be made of the historical data as to whether any of the tests can be eliminated, and/or if additional tests should be added. For illustration, referring again to the above example of a 5000 component lot, from which a sample of 10 components is subjected to 7 quality control tests, it could be that tests
1
-
4
yield 98% of the rejected components, test
5
yields 1.5%, test
6
yields 0.49%, and test
7
yields 0.01%.
Depending upon the level of confidence required, it might be possible to eliminate one or more tests, and the likely candidates for elimination in order would be test
7
, test
6
, and then test
5
. However, merely because a test finds the fewest number of failures, doesn't necessarily mean that it may be eliminated. The selection of which test to eliminate from the quality control testing regimen is not so straight forward.
Furthermore, it may be possible to utilize one or more of the tests on less than all of the selected samples. For example, in the above, perhaps test
7
needs to be conducted on only 5 of the 10 samples.
There is a need in the art for apparatus, methods and products for analyzing test data.
There is another need in the art for apparatus, methods and products for analyzing test data from a multiplicity of tests.
There is even another need in the art for apparatus, methods and products for analyzing test data from a multiplicity of tests to determine the validity of each test of the multiplicity of tests.
These and other needs in the art will become apparent to those of skill in the art upon review of this specification, including its drawings and claims.
SUMMARY OF THE INVENTION
It is an object of the present invention to provide for apparatus, methods and products for analyzing test data.
It is another object of the present invention to provide for apparatus, methods and products for analyzing test data from a multiplicity of tests.
It is even another object of the present invention to provide for analyzing test data from a multiplicity of tests to determine the validity of each test of the multiplicity of tests.
These and other objects of the present invention will become apparent to those of skill in the art upon review of this specification, including its drawings and claims.
According to one embodiment of the present invention, there is provided a method for analyzing a quality control regimen comprising M quality control tests Ti wherein i is=M a finite positive integer≧1, wherein each quality control test T
i
is applied to a population P at a sampling rate R
i
generating test data D
i
. The method generally includes determining for each of the quality control tests T
i
, a C
Pi
and C
PKi
, wherein C
pi
is a measure of the spread of data D
i
, and wherein C
PK
is a measure of the magnitude of how close &mgr;
i
is to the closer of U
i
or L
i
, wherein U
i
, L
i
, and &mgr;
i
, are respectively, the upper limit, lower limit, and mean for D
i
. Optionally, the method further includes adjusting one or more of R
i
, U
i
or L
i
, for at least one of the tests T
i
. In a further embodiment of this embodiment, the method further comprising repeating the determination above. In even a further embodiment of this embodiment, the method includes iteratively repeating the determination and adjusting steps until C
pi
and C
pki
are at desired values.
In still a further embodiment of this embodiment, C
pi
=(U
i
−L
i
)/(2N&sgr;
i
) and C
PKi
={[(lesser of |U
i
−&mgr;
i
| or |L
i
−&mgr;
i
|)]/(N&sgr;
i
/2)}, wherein &sgr;
i
is a standard deviation for D
i
. In yet a further embodiment of this embodiment, the method includes graphically displaying C
Pi
and C
PKi
, with the magnitude of C
pi
represented by a bar graph, with the magnitude of C
PKi
represented as a position of the bar on the graph, and with &mgr;
i
represented as a marking on the bar.
According to another embodiment of the present invention, there is provided an apparatus for analyzing a quality control regimen comprising M quality control tests T
i
wherein i is=M a finite positive integer≧1, wherein each quality control test T
i
is applied to a population P at a sampling rate R
i
generating test data D
i
. Optionally, the apparatus includes a processor provided with instructions that when executed cause the processor to determine for each of the quality control tests T
i
, a C
Pi
and C
PKi
, wherein C
pi
is a measure of the spread of data D
i
, and wherein C
PK
is a measure of the magnitude of how close &mgr;
i
is to the closer of U
i
or L
i
, wherein U
i
, L
i
, and &mgr;
i
, are respectively, the upper limit, lower limit, and mean for D
i
, and also includes an input device for providing to the processor updated values for one or more of R
i
, U
i
or L
i
, for at least one of the tests T
i
. In a further embodiment of the instructions when executed further cause the processor to determine for each of the quality control tests T
i
, updated values for C
Pi
and C
PKi
, based on updated values
Bui Bryan
Haynes and Boone LLP
Pintail Technologies, Inc.
LandOfFree
Method, apparatus and product for evaluating test data does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method, apparatus and product for evaluating test data, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method, apparatus and product for evaluating test data will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3242461