Data processing: measuring – calibrating – or testing – Measurement system in a specific environment – Quality evaluation
Reexamination Certificate
1996-03-12
2002-03-05
Stamber, Eric B. (Department: 2163)
Data processing: measuring, calibrating, or testing
Measurement system in a specific environment
Quality evaluation
C702S081000, C700S108000, C700S109000
Reexamination Certificate
active
06353802
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to reject analysis.
BACKGROUND OF THE INVENTION
One of the major objectives of reject analysis is to ensure that a process is working within or better than the bounds of its historical capability. Constant monitoring of the rejects from a process is the first step to ensure that procedures, especially new ones, are not having an adverse effect on the overall output.
It is well-known to control a process so that it operates within specified boundaries. This can be achieved using statistical process control (SPC) techniques which involve constant monitoring of the process. Such techniques may be univariate wherein a single variable of the process is monitored or multivariate where more than one variable is monitored. Multivariate SPC techniques are particularly well suited to use with complex processes in which a large number of variables are monitored routinely to assess the status of a particular process. Some of the variables may not be independent and the degree to which they are correlated is often unknown, and such processes cannot be assessed adequately with conventional control techniques.
A single parameter known as Hotelling's T
2
(Hotelling, H, (1931), The Generalisation of Student's Ratio,
Ann. Math. Statist
., 2, pages 360-378) can be used successfully as an indicator in multivariate SPC techniques to determine the current status of the process. This parameter utilises all the information contained in the monitored variables as well as accounting for any correlation between them. The state of a process is determined by the magnitude of T
2
, for example, if it exceeds the 95% limit, then the process is behaving in a significantly different way to that of the standard.
The underlying analysis required to deduce the T
2
parameter provides a method of quickly identifying causes of process failure. Corrective action guidelines (CAG) can be developed to facilitate the operation of the system and to provide help for common control failure conditions.
Problem to be Solved by the Invention
Previously known procedures for reject analysis were effectively based on univariate statistical process control (SPC) techniques. However, these techniques are not suitable for use with complex processes where, for each process, a large number of variables are routinely monitored to assess the status of that process. Some of the variables in such a process may not be independent and the degree to which the variables are correlated is often unknown, making it difficult to assess the status of the process.
SUMMARY OF THE INVENTION
It is therefore an object of the present invention to implement multivariate statistical process control (SPC) techniques for reject analysis in order to overcome the problems mentioned above.
In accordance with one aspect of the present invention, there is provided a method of carrying out reject analysis on a repetitive process, the method comprising the steps of:
a) defining a set of variables for the process;
b) sampling data relating to rejects obtained from the process for the defined set of variables;
c) defining a model of the process from the sampled data;
d) applying limits to the model indicative of out-of-control conditions;
e) monitoring the process for out-of-control situations; and
f) taking corrective action to bring the process back into control when the applied limits have been exceeded;
characterised in that the model is defined using principal component analysis in terms of parameters T
2
and Q
res
, where T
2
is derived from the sum of the squares of the scores of each of the principal components of the model and Q
res
is derived from a weighted sum of the squares of the scores of the principal components not included in the model.
If the T
2
parameter exceeds a predetermined limit, the contribution of the scores to that T
2
parameter value is interrogated to determine which score is the primary contributor. The score which forms the primary contributor is interrogated further to assess which of the monitored variables is of significance.
An additional parameter Q
res
may also be determined for the process. If either of the T
2
or Q
res
parameters exceeds predetermined limits, then it indicates a significant change compared with the reference system.
T
2
and Q
res
monitor different out-of-control behaviour, T
2
assessing systematic variability within the model and Q
res
the systematic non-random variability not captured by the model.
The term ‘repetitive process’ means a routine process which is repeated according to requirement for that process. For example, processing of X-ray films in a radiology department is a routine process which is repeated as required.
Advantageous Effect of the Invention
By using multivariate SPC techniques, the sensitivity of detecting out-of-control conditions for reject analysis is increased in comparison with existing manual and labour intensive methods.
In accordance with the present invention, the method of reject analysis described herein provides a key parameter on which to base a quality assurance strategy. As a result, the method is also applicable in other applications, for example, photofinishing.
It also provides a sound statistical foundation on which to formulate any quality improvement strategy, and a simple parameter is derived for used in everyday operations.
A potential benefit of using Hotelling's T
2
parameter is that it additionally yields vital information which can be used to correct any control failure problem with efficacy.
In particular T
2
determines, the aim and limits for rejects from a process in terms of monitored variables or classification categories. An immediate assessment of any analysis period relative to a reference position can be determined. Moreover, the method also allows the causes of increased or decreased reject rates to be quickly established and then measures, can be implemented either to correct problems or new algorithms generated to reflect the improved working position.
Furthermore, the use of the method of the present invention ensures that other measured variables or categories found to impact the reject performance characteristics of a process at a later stage can be included as an extension to that method at any time.
REFERENCES:
patent: 5121337 (1992-06-01), Brown
patent: 5406502 (1995-04-01), Haramaty et al.
patent: 5442562 (1995-08-01), Hopkins et al.
patent: 5446681 (1995-08-01), Gethner et al.
patent: 5592402 (1997-01-01), Beebe et al.
Martin et al., “Process Performance Monitoring Using Multivariate Statistical Process Control”, IEE Proceedings on Control Theory and Applications, vol. 143, No. 2, Mar. 1996, pp. 132-144.*
Babiloni et al., “Statistical Analysis of Topographic Maps of Short-Latency Somatosensory Evoked Potentials in Normal and Parkinsonian Subjects”, IEEE Trans. on Biomedical Engineering, vol. 41, No. 7, Jul. 1994, pp. 617-624.*
Chee K. Chow, “Projection of Circuit Performance Distributions by Multivariate Statistics”, IEEE Trans. on Semiconductor Manufacturing, vol. 2, No. 2, May 1989, pp. 60-65.*
Ignova et al., “Neural Networks and Nonlinear SPC”, Proceedings of the Third IEEE Conference on Control Applications, Aug. 1994, pp. 1271-1276.*
Piovoso et al., “Process Data Chemometrics”, IEEE Instrumentation and Measurement Technology Conference, May 1991, pp. 608-612A.*
Efthimaidu et al., “Online Statistical Process Control of Chemical Processes”, International Conference on Control, Mar. 1991, pp. 700-705.*
H. Hotelling, The Annals of Mathematical Statistics,The Generalization of Student's Ratio, 1931, vol. 2, pp. 360-378.
D.C. Montgomery, Introduction to Statistical Process Control, Wiley, 1991, pp. 109-127.
G.W. Snedecor and W.G. Cochran, Statistical Methods, 1980, 7th Ed., Iowa State University Press, pp. 220-223.
P. Miller, R.E. Swanson and C.F. Heckler, Contribution Plots: The Missing Link in Multivariate Quality Control, 37th Annual Fall Conference, 1993, ASQC, Rochester, New York, USA.
Barbur Vicki Ann
Green Andrew
Close Thomas H.
Eastman Kodak Company
Stamber Eric B.
LandOfFree
Reject analysis does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Reject analysis, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Reject analysis will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2840203