System and method for automatically predicting the timing and costs of service events in a life cycle of a product

• US 6,832,205 B1
• Filed: 06/30/2000
• Issued: 12/14/2004
• Est. Priority Date: 06/30/2000
• Status: Active Grant

• Current Assignee: General Electric Company
• Original Assignee: General Electric Company
• Inventors: Aragones, James Kenneth, Stein, Jeffrey William, Aragones, Amy Victoria, Tucker, William Talbert
• Primary Examiner(s): Jeanty, Romain
• Application Number:
  US09/608,628
• Time in Patent Office: 1,628 Days
• Field of Search: 705/10, 705/11, 702/90, 702/33, 702/179, 703/6, 703/7, 703/8
• US Class Current: 705/7.25
• CPC Class Codes: G06Q 10/06 Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
• CPC Class Codes: G06Q 10/06315 Needs-based resource requirements planning or analysis
• CPC Class Codes: G06Q 10/06375 Prediction of business process outcome or impact based on a proposed change
• CPC Class Codes: G06Q 30/0201 Market modelling; Market analysis; Collecting market data

Abstract

A system and method for automatically predicting timing and costs of future service events in a life cycle of a product. A database contains a plurality of service information and performance information for the product. A statistical analyzer automatically analyzes the plurality of processed service information to determine a plurality of compartment failure information. A performance deterioration rate analyzer automatically analyzes the performance deterioration rate of the product from the plurality of service information and performance information. A simulator, automatically simulates a distribution of future service events in the life cycle of the product and predicts the costs of the service events according to the plurality of compartment failure information and the performance deterioration rate analysis.

Claims

1. An automated system for predicting the timing and costs of future service events in a life cycle of a product formed from a plurality of compartments, comprising:

a database that contains a plurality of service information and a plurality of performance information for the product;

a statistical analyzer that automatically analyzes the plurality of service information to determine a plurality of compartment failure information comprising compartment failure variables and compartment time-to-failure coefficients, wherein the statistical analyzer uses the plurality of compartment failure information to determine which compartment failure variables influence the timing of future service events and estimate time-to-failure distributions for the plurality of compartments;

a performance deterioration rate analyzer that automatically analyzes performance deterioration rate of the product from the plurality of service information and performance information, wherein the performance deterioration rate analyzer comprises a statistical analysis script that relates a subset of compartments of the product according to time, wherein the statistical analysis script generates an estimated deterioration rate curve for the subset of compartments of the product, wherein the performance deterioration rate analyzer further comprises a transformer that transforms each estimated deterioration rate curve for a compartment to a performance life distribution; and

a simulator that automatically simulates a distribution of future service events in the life cycle of the product and predicts the costs of the service events according to the time-to-failure distributions and performance life distributions.

2. The system according to claim 1, wherein the database comprises a service database and a performance historical database.

3. The system according to claim 1, wherein the plurality of service information comprises compartment definitions, repair history and service factors.

4. The system according to claim 1, wherein the plurality of performance information comprises performance characteristic values, initial data levels after servicing, current data levels, dates at which the product is serviced, and variables that affect the servicing of a subset of the plurality of compartments.

5. The system according to claim 1, further comprising a preprocessor that automatically processes the plurality of service information into a predetermined format.

6. The system according to claim 5, wherein the preprocessor generates a plurality of data files according to the plurality of service information.

7. The system according to claim 1, wherein the statistical analyzer uses the estimated time-to-failure distributions to determine a Weibull distribution for a subset of the plurality of compartments defined for the product.

8. The system according to claim 1, wherein the statistical analyzer comprises a service analysis script that executes a plurality of statistical procedures.

9. The system according to claim 8, wherein the plurality of statistical procedures comprise a multivariate regression and/or a correlation analysis.

10. The system according to claim 8, wherein the service analysis script generates a plurality of statistical diagnostic information.

11. The system according to claim 10, wherein the plurality of statistical diagnostic information comprises goodness-of-fit metrics and collinearity diagnostics.

12. The system according to claim 8, wherein the service analysis script generates a plurality of residual plots.

13. The system according to claim 1, wherein the statistical analyzer comprises a validation script.

14. The system according to claim 13, wherein the validation script is applied to a plurality of case studies set up for the product.

15. The system according to claim 1, wherein the simulator uses the performance life distributions to determine a Weibull distribution for a subset of the plurality of compartments defined for the product.

16. The system according to claim 1, wherein the simulator forecasts a service plan for the future service events that comprises the time for scheduling the service events and the costs associated therewith.

17. An automated system for predicting the timing and costs of future service events in a life cycle of a product formed from a plurality of compartments, comprising:

means for storing a plurality of service information and a plurality of performance information for the product;

means for automatically analyzing the plurality of service information to determine a plurality of compartment failure information comprising compartment failure variables and compartment time-to-failure coefficients, wherein the analyzing means uses the plurality of compartment failure information to determine which compartment failure variables influence the timing of future service events and estimate time-to-failure distributions for the compartments;

means for automatically performing a deterioration rate analysis that determines performance deterioration rate of the product from the plurality of service information and performance information, wherein the performing means comprises a statistical analysis script that relates a subset of the plurality of compartments of the product according to time, wherein the statistical analysis script generates an estimated deterioration rate curve for a subset of the plurality of compartments of the product, wherein the performing means further comprises means for transforming each estimate deterioration rate curve for a compartment to a performance life distribution; and

means for automatically simulating a distribution of future service events in the life cycle of the product and predicting the costs of the service events according to the time-to-failure distributions and performance life distributions.

18. Tho system according to claim 17, wherein the plurality of service information comprises compartment definitions, repair history and service factors.

19. The system according to claim 17, wherein the plurality of performance information comprises performance characteristic values, initial data levels after servicing, current data levels, dates at which the product is serviced, and variables that affect the servicing of a subset of the plurality of compartments of the product.

20. The system according to claim 17, further comprising means for automatically preprocessing the plurality of service information into a predetermined format.

21. The system according to claim 20, wherein the preprocessing means generates a plurality of data files according to the plurality of service information.

22. The system according to claim 17, wherein the analyzing means uses the estimated time-to-failure distributions to determine a Weibull distribution for a subset of the plurality of compartment defined for the product.

23. The system according to claim 17, wherein the analyzing means comprises a service analysis script that executes a plurality of statistical procedures.

24. The system according to claim 23, wherein the plurality of statistical procedures comprise a multivariate regression and/or a correlation analysis.

25. The system according to claim 23, wherein the service analysis script generates a plurality of statistical diagnostic information.

26. The system according to claim 25, wherein the plurality of statistical diagnostic information comprises goodness-of-fit metrics and collinearity diagnostics.

27. The system according to claim 23, wherein the service analysis script generates a plurality of residual plots.

28. The system according to claim 17, wherein the analyzing means comprises a validation script.

29. The system according to claim 28, wherein the validation script is applied to a plurality of case studies set up for the product.

30. The system according to claim 17, wherein the simulating means uses the performance life distribution to determine a Weibull distribution for a subset of the plurality of compartments defined for the product.

31. The system according to claim 17, wherein the simulating means forecasts a service plan for the future service events that comprises the time for scheduling the service events and the costs associated therewith.

32. A method for automatically predicting the timing and costs of future service events in a life cycle of a product formed from a plurality of compartments, comprising:

storing a plurality of service information and a plurality of performance information for the product;

automatically analyzing the plurality of service information to determine a plurality of compartment failure information comprising compartment failure variables and compartment time-to-failure coefficients, wherein the analyzing uses the plurality of compartment failure information to determine which compartment failure variables influence the timing of future service events and estimate time-to-failure distributions for the plurality of compartments;

automatically performing a deterioration rate analysis that determines performance deterioration rate of the product from the plurality of service information and performance information, wherein the performing comprises using a statistical analysis script that relates a subset of the plurality of compartments of the product according to time, wherein the statistical analysis script generates an estimated deterioration rate curve for a subset of the plurality of compartments of the product, wherein the performing a deterioration rate analysis further comprises transforming each estimated deterioration rate curve for a compartment to a performance life distribution; and

automatically simulating a distribution of future service events in the life cycle of the product and predicting the costs of the service events according to the time-to-failure distributions and performance life distributions.

33. The method according to claim 32, wherein the plurality of service information comprises compartment definitions, repair history and service factors.

34. The method according to claim 32, wherein the plurality of performance information comprises performance characteristic values, initial data levels after servicing, current data levels, dates at which the product is serviced, and variables that affect the servicing of a subset of the plurality of compartments of the product.

35. The method according to claim 32, further comprising automatically preprocessing the plurality of service information into a predetermined format.

36. The method according to claim 35, wherein the preprocessing generates a plurality of data files according to the plurality of service information.

37. The method according to claim 32, wherein the analyzing uses the estimated time-to-failure distributions to determine a Weibull distribution for a subset of the plurality of compartments.

38. The method according to claim 32, wherein the analyzing comprises using a service analysis script that executes a plurality of statistical procedures.

39. The method according to claim 38, wherein the plurality of statistical procedures comprise a multivariate regression and/or a correlation analysis.

40. The method according to claim 39, wherein the service analysis script generates a plurality of statistical diagnostic information.

41. The method according to claim 40, wherein the plurality of statistical diagnostic information comprises goodness-of-fit metrics and collinearity diagnostics.

42. The method according to claim 38, wherein the service analysis script generates a plurality of residual plots.

43. The method according to claim 32, wherein the analyzing comprises a validation script.

44. The method according to claim 43, wherein the validation script is applied to a plurality of case studies set tip for the product.

45. The method according to claim 32, wherein the simulating uses the performance life distributions to determine a Weibull distribution for a subset of the plurality of compartments.

46. The method according to claim 32, wherein the simulating forecasts a service plan for the future service events that comprises the time for scheduling the service events and the costs associated therewith.

47. A computer-readable medium storing computer instructions for instructing a computer system to automatically predict the timing and costs of future service events in a life cycle of a product formed from a plurality of compartments, the computer instructions comprising:

storing a plurality of service information and a plurality of performance information for the product;

automatically analyzing the plurality of service information to determine a plurality of compartment failure information comprising compartment failure variables and compartment time-to-failure coefficients, wherein the analyzing instructions uses the plurality of compartment failure information to determine which compartment failure variables influence the timing of future service events and estimate time-to-failure distributions for the plurality of compartments;

automatically performing a deterioration rate analysis that determines performance deterioration rate of the product from the plurality of service information and performance information, wherein the performing instructions comprise a statistical analysis script that relates a subset of the plurality of compartments of the product according to time, wherein the statistical analysis script a generates an estimated deterioration rate curve for a subset of the plurality of compartments of the product, wherein the performing instructions further comprise transforming instructions that transform each estimated deterioration rate curve to a performance life distribution; and

automatically simulating a distribution of future service events in the life cycle of the product and predicting the costs of the service events according to the time-to-failure distribution and performance life distributions.

48. The computer-readable medium according to claim 47, wherein the plurality of service information comprises compartment definitions, repair history and service factors.

49. The computer-readable medium according to claim 47, wherein the plurality of performance information comprises performance characteristic values, initial data levels after servicing, current data levels, dates at which the product is serviced, and variables that affect the servicing of a subset of the plurality of compartments of the product.

50. The computer-readable medium according to claim 47, further comprising preprocessing instructions that automatically preprocess the plurality of service information into a predetermined format.

51. The computer-readable medium according to claim 50, wherein the preprocessing instructions generates a plurality of data files according to the plurality of service information.

52. The computer-readable medium according to claim 47, wherein the analyzing instructions use the estimated time-to-failure distributions to determine a Weibull distribution for a subset of the plurality of compartments.

53. The computer-readable medium according to claim 47, wherein the analyzing instructions comprises instructions for using a service analysis script that executes a plurality of statistical procedures.

54. The computer-readable medium according to claim 53, wherein the plurality of statistical procedures comprise a multivariate regression and/or a correlation analysis.

55. The compute-readable medium according to claim 54, wherein the service analysis script generates a plurality of statistical diagnostic information.

56. The computer-readable medium according to claim 55, wherein the plurality of statistical diagnostic information comprises goodness-of-fit metrics and collinearity diagnostics.

57. The computer-readable medium according to claim 53, wherein the service analysis script generates a plurality of residual plots.

58. The computer-readable medium according to claim 47, wherein the analyzing instructions comprise a validation script.

59. The computer-readable medium according to claim 58, wherein the validation script is applied to a plurality of case studies set up for the product.

60. The computer-readable medium according to claim 47, wherein the simulating instructions use the performance life distribution to determine a Weibull distribution for a subset of the plurality of compartments.

61. The computer-readable medium according to claim 47, wherein the simulating instructions forecasts a service plan for the future service events that comprises the time for scheduling the service events and the costs associated therewith.