Methods and apparatus for data analysis

• US 8,041,541 B2
• Filed: 03/27/2007
• Issued: 10/18/2011
• Est. Priority Date: 05/24/2001
• Status: Expired due to Fees

• Current Assignee: In-Depth Test LLC (Acacia Research Corporation)
• Original Assignee: Test Advantage Incorporated (Boston Semi Equipment, LLC)
• Inventors: Martin, Emilio Miguelanez, Buxton, Paul M., Zalzala, Ali M. S., Tabor, Eric Paul
• Primary Examiner(s): KHUU, HIEN DIEU THI
• Application Number:
  US11/692,021
• Publication Number:
  US 20080021677A1
• Time in Patent Office: 1,666 Days
• Field of Search: 702/34, 702/35, 702/108, 702117-119, 702/123, 702182-185, 324/73.1, 324/500, 714/25, 714/33, 714/100
• US Class Current: 702/183
• CPC Class Codes: G01R 31/2894 Aspects of quality control [QC] G01R31/31718 takes precedence; program control for QC G05B19/41875
• CPC Class Codes: G05B 19/41875 characterised by quality surveillance of production
• CPC Class Codes: Y02P 90/02 Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]

Abstract

A method and apparatus for data analysis according to various aspects of the present invention is configured to automatically identify a characteristic of a fabrication process for components based on test data for the components.

Claims

1. A test system, comprising:

a tester configured to test a set of components and generate test data for the set of components, wherein the components are fabricated in accordance with a fabrication process; and

a diagnostic system configured to receive the test data from the tester, automatically analyze the test data, and identify a characteristic of the fabrication process for the components according to the analysis, wherein the diagnostic system comprises a pattern recognition system configured to;

recognize a pattern in the test data;

compare the recognized pattern to a known pattern associated with the characteristic; and

automatically learn an additional pattern associated with a new characteristic based on the recognized pattern, wherein automatically learn the additional pattern associated with the new characteristic does not require the pattern recognition system to retrain.

2. A test system according to claim 1, wherein the test data comprises at least one of electronic wafer sort data, data derived from electronic wafer sort data, electrical test data, bin map data, and outlier data.

3. A test system according to claim 1, wherein the diagnostic system is configured to provide a corrective action suggestion based on the identified characteristic.

4. A test system according to claim 1, wherein the pattern recognition system comprises a classifier configured to classify the recognized pattern according to a known pattern.

5. A test system according to claim 4, wherein the classifier comprises a neural network.

6. A test system according to claim 5, wherein the neural network comprises a radial basis function network.

7. A test system according to claim 1, wherein the pattern recognition system includes a feature extractor configured to extract a feature from the test data associated with the pattern.

8. A test system according to claim 7, wherein the feature extractor calculates at least one of a mass, a centroid, a geometric moment, and a moment of Hu based on the test data.

9. A test system according to claim 7, wherein the feature extractor is configured to extract at least two features from the test data, and wherein the pattern recognition system further comprises a feature selector configured to select fewer than all of the features for analysis.

10. A test system according to claim 9, wherein the feature selector operates in conjunction with a genetic algorithm.

11. A test data analysis system for analyzing test data for a set of components fabricated and tested using a fabrication process, comprising:

a memory for storing the test data; and

a diagnostic system having access to the memory and configured to analyze the test data, and identify a characteristic of the fabrication process based on the test data, wherein the diagnostic system comprises a pattern recognition system configured to;

recognize a pattern in the test data;

compare the recognized pattern to a known pattern associated with the characteristic; and

automatically learn an additional pattern associated with a new characteristic based on the recognized pattern, wherein automatically learn the additional pattern associated with the new characteristic does not require the pattern recognition system to retrain.

12. A test data analysis system according to claim 11, wherein the test data comprises at least one of electronic wafer sort data, data derived from electronic wafer sort data, electrical test data, bin map data and outlier data.

13. A test data analysis system according to claim 11, wherein the diagnostic system is configured to provide a corrective action suggestion based on the identified characteristic.

14. A test data analysis system according to claim 11, wherein the pattern recognition system comprises a classifier configured to classify the recognized pattern according to a known pattern.

15. A test data analysis system according to claim 14, wherein the classifier comprises a neural network.

16. A test data analysis system according to claim 15, wherein the neural network comprises a radial basis function network.

17. A test data analysis system according to claim 11, wherein the pattern recognition system includes a feature extractor configured to extract a feature from the test data associated with the pattern.

18. A test data analysis system according to claim 17, wherein the feature extractor calculates at least one of a mass, a centroid, a geometric moment, and a moment of Hu based on the test data.

19. A test data analysis system according to claim 17, wherein the feature extractor is configured to extract at least two features from the test data, and wherein the pattern recognition system further comprises a feature selector configured to select fewer than all of the features for analysis.

20. A test data analysis system according to claim 19, wherein the feature selector operates in conjunction with a genetic algorithm.

21. A computer-implemented method for testing components fabricated and tested according to a fabrication process, comprising:

obtaining test data for the components from a computer memory; and

analyzing the test data and automatically identifying a characteristic of the fabrication process based on the test data by the computer, wherein automatically identifying the characteristic comprises;

recognizing a pattern in the test data;

comparing the recognized pattern to a known pattern associated with the characteristic; and

automatically learning an additional pattern associated with a new characteristic based on the recognized pattern, wherein automatically learning the additional pattern associated with the new characteristic does not require retraining.

22. A computer-implemented method for testing components according to claim 21, wherein the test data comprises at least one of electronic wafer sort data, data derived from electronic wafer sort data, electrical test data, bin map data, and outlier data.

23. A computer-implemented method for testing components according to claim 21, further comprising providing a corrective action suggestion based on the identified characteristic.

24. A computer-implemented method for testing components according to claim 21, wherein automatically identifying the characteristic comprises classifying the recognized pattern according to a known pattern.

25. A computer-implemented method for testing components according to claim 24, wherein classifying the recognized pattern is performed by a neural network.

26. A computer-implemented method for testing components according to claim 25, wherein the neural network comprises a radial basis function network.

27. A computer-implemented method for testing components according to claim 21, wherein automatically identifying the characteristic comprises extracting a feature from the test data associated with the recognized pattern.

28. A computer-implemented method for testing components according to claim 27, wherein the feature comprises at least one of a mass, a centroid, a geometric moment, and a moment of Hu based on the test data.

29. A computer-implemented method for testing components according to claim 27, wherein automatically identifying the characteristic further comprises selecting the feature from multiple features for analysis.

30. A non-transitory computer readable medium containing computer instructions stored therein for causing a computer processor to perform a method for analyzing test data comprising:

obtaining test data for a set of components; and

analyzing the test data and automatically identifying a characteristic of the fabrication process based on the test data, wherein automatically identifying the characteristic comprises;

recognizing a pattern in the test data;

comparing the recognized pattern to a known pattern associated with the characteristic; and

automatically learning an additional pattern associated with a new characteristic based on the recognized pattern, wherein automatically learning the additional pattern associated with the new characteristic does not require retraining.

31. A non-transitory computer readable medium according to claim 30, wherein the test data comprises at least one of electronic wafer sort data, data derived from electronic wafer sort data, electrical test data, bin map data, and outlier data.

32. A non-transitory computer readable medium according to claim 30, the method for analyzing further comprising providing a corrective action suggestion based on the identified characteristic.

33. A non-transitory computer readable medium according to claim 30, wherein automatically identifying the characteristic comprises classifying the recognized pattern according to a known pattern.

34. A non-transitory computer readable medium according to claim 33, wherein classifying the recognized pattern is performed by a neural network.

35. A non-transitory computer readable medium according to claim 34, wherein the neural network comprises a radial basis function network.

36. A non-transitory computer readable medium according to claim 30, wherein automatically identifying the characteristic comprises extracting a feature from the test data associated with the recognized pattern.

37. A non-transitory computer readable medium according to claim 36, wherein the feature comprises at least one of a mass, a centroid, a geometric moment, and a moment of Flu based on the test data.

38. A non-transitory computer readable medium according to claim. 36, wherein automatically identifying the characteristic further comprises selecting the feature from multiple features for analysis.