Intelligent agent system and method for evaluating data integrity in process information databases

• US 6,968,250 B2
• Filed: 11/27/2002
• Issued: 11/22/2005
• Est. Priority Date: 12/28/2001
• Status: Active Grant

• Current Assignee: Callahan Cellular LLC (Intellectual Ventures LLC)
• Original Assignee: Kimberly-Clark Worldwide Incorporated (Kimberly-Clark Corporation)
• Inventors: Reade, Walter Caswell, Barron Barber, Douglas Gordon, Fuller, Paul D., Klaips, Melissa S., Markham, Charles Earl, Pokorny, Michael Roy
• Primary Examiner(s): Gandhi, Jayprakash N.
• Application Number:
  US10/306,795
• Publication Number:
  US 20030149373A1
• Time in Patent Office: 1,091 Days
• Field of Search: 700/32, 700/95, 700/108, 700/109, 700/132, 702/84
• US Class Current: 700/109
• CPC Class Codes: B23Q 35/12 involving electrical means programme recording for copying purposes in a separate apparatus G05, G11
• CPC Class Codes: G05B 13/0285 using neural networks and fuzzy logic
• CPC Class Codes: G06N 5/048 Fuzzy inferencing
• CPC Class Codes: G06Q 10/06 Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
• CPC Class Codes: G06Q 10/10 Office automation; Time management

Abstract

A data integrity module and method for evaluating data in a process information database. A neural network generates statistical patterns for specifying patterns for the data being evaluated. A fuzzy expert rules base specifies rules for evaluating the data. A processor, responsive to the rules base and the statistical patterns, identifies suspect data in the process information database by evaluating the data according to the rules base and the statistical patterns. A modification system modifies the suspect data in the process information database.

Claims

1. A data integrity system for evaluating data in a process information database including data generated by a manufacturing process, said system comprising:

an evaluation resource cooperatively associated with reference information, wherein the evaluation resource comprises a neural network and wherein the reference information comprises statistical patterns generated by the neural network for specifying patterns for the data being evaluated;

a processor, responsive to the reference information, for accessing the data in the process information database and for evaluating the accessed data according to the reference information, the processor identifying suspect data in the evaluated data; and

a modification subsystem responsive to the processor for modifying at least some of the suspect data in the process information database.

2. The system of claim 1 wherein the reference information comprises a fuzzy expert rules base specifying rules for evaluating the data.

3. The system of claim 1 further comprising a storage module for storing a clear data set which does not include suspect data, and wherein the neural network generates the statistical patterns based on the clear data set.

4. The system of claim 1 wherein the reference information further comprises a fuzzy expert rules base specifying rules for evaluating the data, and further comprising a marking subsystem for marking evaluated data as one or more of the following:

suspect when its accuracy is inconsistent with the reference information, auto-corrected when suspect data has been corrected without user intervention, user-corrected when the suspect data has been corrected by the user, and/or checked when its accuracy is consistent with the reference information.

5. The system of claim 4 wherein the neural network performs a higher-order evaluation relating to a historical analysis of data trends and/or analyzes whether or not data is present in the database and/or whether data is above or below a certain value.

6. A data integrity system for evaluating data in a process information database including data generated by a manufacturing process, said system comprising:

an evaluation resource cooperatively associated with reference information, wherein the reference information comprises a rules base specifying rules for evaluating the data;

a processor, responsive to the reference information, for accessing the data in the process information database and for evaluating the accessed data according to the reference information, the processor identifying suspect data in the evaluated data; and

a modification subsystem responsive to the processor for modifying at least some of the suspect data in the process information database.

7. A data integrity module for evaluating data in a process information database including data generated by a manufacturing process, said module comprising:

a neural network generating statistical patterns for specifying patterns for the data being evaluated;

a rules base specifying rules for evaluating the data;

a processor1 responsive to the rules base and the statistical patterns, for identifying suspect data in the process information database by evaluating the data according to the rules base and the statistical patterns; and

modification system for modifying the suspect data in the process information database.

8. The module of claim 7 wherein the modification system comprises a modification subsystem responsive to the processor for modifying at least some of the suspect data in the process information database and a user access subsystem for permitting a user to modify certain of the identified suspect data.

9. The module of claim 7 further comprising a storage module for storing a clear data set which does not include suspect data and wherein the neural network generates the statistical patterns based on the clear data set.

10. The module of claim 7 further comprising a marking subsystem for marking evaluated data as one or more of the following:

suspect when its accuracy is inconsistent with the rules base and the statistical patterns;

auto-corrected when suspect data has been corrected without user intervention;

user-corrected when the suspect data has been corrected by the user; and

checked when its accuracy is consistent with the rules base and the statistical patterns.

11. The module of claim 7 wherein the process information database comprises one or more of the following:

a raw materials database, a waste and delay database, a productivity database, a quality database and a machine process database.

12. The module of claim 7 wherein the manufacturing process produces a product including one or more of the following products or parts therefor:

pharmaceuticals, automobiles, food and beverage, pulp and paper, injection molded items, electronics, and printed items.

13. The module of claim 7 wherein the neural network performs a higher-order evaluation relating to a historical analysis of data trends and/or analyzes whether or not data is present in the database and/or whether data is above or below a certain value.

14. The module of claim 7 wherein the rules base further comprises a fuzzy expert rules base specifying rules for evaluating the data.

15. In a system for manufacturing absorbent garments from raw material by a process implemented by a product line, wherein the system has a process information database of data relating to the process, the improvement comprising a data integrity module for evaluating the data in the process information database including data generated by a manufacturing process, said module comprising:

a neural network generating statistical patterns for specifying patterns for the data being evaluated;

a rules base specifying rules for evaluating the data;

a processor, responsive to the rules base and the statistical patterns, for identifying suspect data in the process information database by evaluating the data according to the rules base and the statistical patterns; and

a modification system for modifying the suspect data in the process information database.

16. The system of claim 15 wherein the modification system comprises a modification subsystem responsive to the processor for modifying at least some of the suspect data in the process information database and a user access subsystem for permitting a user to modify certain of the identified suspect data.

17. The system of claim 15 further comprising a storage module for storing a clear data set which does not include suspect data and wherein the neural network generates the statistical patterns based on the clear data set.

18. The system of claim 15 further comprising a marking subsystem for marking evaluated data as one or more of the following:

suspect when its accuracy is inconsistent with the rules base and the statistical patterns;

auto-corrected when suspect data has been corrected without user intervention;

user-corrected when the suspect data has been corrected by the user; and

/or checked when its accuracy is consistent with the rules base and the statistical patterns.

19. The system of claim 18 wherein the process information database comprises one or more of the following:

a raw materials database, a waste and delay database, a productivity database, a quality database and a machine process database.

20. The system of claim 15 wherein the neural network performs a higher-order evaluation relating to a historical analysis of data trends and/or analyzes whether or not data is present in the database and/or whether data is above or below a certain value.

21. The system of claim 15 wherein the manufacturing process produces a product including one or more of the following products or parts therefor:

pharmaceuticals, automobiles, food and beverage, pulp and paper, injection molded items, electronics, and printed items.

22. The module of claim 15 wherein the rules base further comprises a fuzzy expert rules base specifying rules for evaluating the data.

23. A method for evaluating data integrity in a process information database including data generated by a manufacturing process, said method comprising:

providing reference information, wherein the reference information comprises statistical patterns generated by a neural network for specifying patterns for the data being evaluated;

accessing the data in the process information database;

evaluating the accessed data according to the reference information;

identifying suspect data in the evaluated data; and

modifying at least some of the suspect data in the process information database.

24. The method of claim 23 wherein the reference information comprises a rules base specifying rules for evaluating the data.

25. The method of claim 23 further comprising storing a clear data set which does not include suspect data and wherein the neural network generates the statistical patterns based on the clear data set.

26. The method of claim 23 wherein the neural network performs a higher-order evaluation relating to a historical analysis of data trends and/or analyzes whether or not data is present in the database and/or whether data is above or below a certain value.

27. A method for evaluating data integrity in a process information database including data generated by a manufacturing process, said method comprising:

providing reference information, wherein the reference information comprises a rules base specifying rules for evaluating the data;

accessing the data in the process information database;

evaluating the accessed data according to the reference information;

identifying suspect data in the evaluated data; and

modifying at least some of the suspect data in the process information database.

28. A method for evaluating data integrity in a process information database including data generated by a manufacturing process, said method comprising:

providing reference information;

accessing the data in the process information database;

evaluating the accessed data according to the reference information;

identifying suspect data in the evaluated data;

modifying at least some of the suspect data in the process information database; and

permitting a user to modify certain of the identified suspect data.

29. A method for evaluating data integrity in a process information database including data generated by a manufacturing process, said method comprising:

providing reference information;

accessing the data in the process information database;

evaluating the accessed data according to the reference information;

identifying suspect data in the evaluated data;

modifying at least some of the suspect data in the process information database; and

marking evaluated data as one or more of the following;

suspect when its accuracy is inconsistent with the reference information, auto-corrected when suspect data has been corrected without user interventiont, user-corrected when the suspect data has been corrected by the user, and/or checked when its accuracy is consistent with the reference information.

30. The method of claim 29 wherein the resource comprises a neural network and wherein the reference information comprises statistical patterns generated by a neural network for specifying patterns for the data being evaluated and wherein the reference information further comprises a fuzzy expert rules base specifying rules for evaluating the data.

31. The method of claim 30 wherein the neural network performs a higher-order evaluation relating to a historical analysis of data trends and/or analyzes whether or not data is present in the database and/or whether data is above or below a certain value.

32. A method for evaluating data integrity in a process information database including data generated by a manufacturing process, said method comprising:

providing reference information;

accessing the data in the process information database;

evaluating the accessed data according to the reference information;

identifying suspect data in the evaluated data; and

modifying at least some of the suspect data in the process information database, wherein the process information database comprises one or more of the following;

a raw materials database, a waste and delay database, a productivity database, a quality database and a machine process database.