Intelligent system for detection of process status, process fault and preventive maintenance

• US 20060259198A1
• Filed: 05/26/2006
• Published: 11/16/2006
• Est. Priority Date: 11/26/2003
• Status: Abandoned Application

• Current Assignee: Tokyo Electron Limited
• Original Assignee: Tokyo Electron Limited
• Inventors: Brcka, Jozef, Delp, Deana, Moroz, Paul, Grapperhaus, Michael
• Application Number:
  US11/441,050
• Publication Number:
  US 20060259198A1
• Time in Patent Office: Days
• US Class Current: 700/246
• CPC Class Codes: G05B 19/4184 characterised by fault tolerance, reliability of production system
• CPC Class Codes: G05B 2219/24019 Computer assisted maintenance
• CPC Class Codes: G05B 2219/31357 Observer based fault detection, use model
• CPC Class Codes: G05B 2219/32234 Maintenance planning
• CPC Class Codes: G05B 2219/33296 ANN for diagnostic, monitoring
• CPC Class Codes: G05B 2219/45031 Manufacturing semiconductor wafers
• CPC Class Codes: G05B 23/0254 based on a quantitative model, e.g. mathematical relationships between inputs and outputs; functions: observer, Kalman filter, residual calculation, Neural Networks
• CPC Class Codes: Y02P 90/02 Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
• CPC Class Codes: Y02P 90/80 Management or planning

Abstract

Embodiments of an intelligent modeling method and system monitor and perform analysis of semiconductor processing equipment as well as predict future states of that equipment based on the analysis, predict failures of the semiconductor processing equipment and/or determine equipment maintenance schedules.

Claims

1. A method for determining a future state of semiconductor processing equipment, the method comprising:

training a neural network model with semiconductor processing equipment process measurements to predict failure of the semiconductor processing equipment.

2. The method of claim 1, further comprising updating the model by applying at least one reward value calculated based on past model accuracy in predicting failure of the semiconductor processing equipment.

3. The method of claim 1, further comprising formulating a maintenance schedule for the semiconductor processing equipment based at least in part on the model.

4. The method of claim 3, further comprising updating of the model by applying at least one reward value calculated based on past model accuracy in predicting maintenance requirements.

5. The method of claim 3, wherein the maintenance schedule is formulated using an input data set produced by Principle Component Analysis.

6. The method of claim 1, wherein semiconductor processing equipment processing measurements include at least one of a time of last cleaning of the semiconductor processing equipment, a time of last maintenance of the semiconductor processing equipment, a time of a last failure of the semiconductor equipment and a preventative maintenance schedule for the semiconductor processing equipment.

7. The method of claim 1, wherein the neural network outputs data including at least one of a suggested maintenance schedule for the semiconductor processing equipment, a suggested cleaning schedule for the semiconductor processing equipment and a predicted time for a next failure of the semiconductor processing equipment.

8. The method of claim 1, further comprising obtaining data indicating the semiconductor equipment process measurements used for predicting the failure of the semiconductor equipment.

9. The method of claim 1, wherein the processing measurements include at least one of chamber temperature, gas mixture and applied radio frequency power.

10. The method of claim 1, wherein the model is multivariate.

11. The method of claim 10, wherein the model analyzes data sets organized as, a data matrix, a correlation matrix, a variance-covariance matrix, a sum-of-squares, a cross-products matrix, or a sequence of residuals.

12. The method of claim 1, wherein the model is dynamic and the model is updated as processing measurement data is produced by the semiconductor processing equipment.

13. The method of claim 1, wherein the model is formulated using offline simulations to completely model the semiconductor processing equipment.

14. The method of claim 13, further comprising testing the model for reliability and validating the model by comparing previously predicted states with processing measurements produced by the semiconductor processing equipment.

15. The method of claim 1, further comprising providing an icon driven user interface as a front end that allows a user to retrieve data and obtain processing measurements from the semiconductor processing equipment.

16. A system for determining a future state of semiconductor processing equipment, the system comprising:

a neural network model trained with semiconductor processing equipment process measurements to predict failure of semiconductor processing equipment.

17. The system of claim 16, wherein the model is updatable by applying at least one reward value calculated based on past model accuracy in predicting failure of the semiconductor processing equipment.

18. The system of claim 16, wherein semiconductor processing equipment processing measurements include at least one of a time of last cleaning of the semiconductor processing equipment, a time of last maintenance of the semiconductor processing equipment, a time of a last failure of the semiconductor equipment and a preventative maintenance schedule for the semiconductor processing equipment.

19. The system of claim 16, wherein the model formulates a maintenance schedule for the semiconductor processing equipment based at least in part on the model.

20. The system of claim 19, wherein the model is updatable by applying at least one reward value calculated based on past model accuracy in predicting maintenance requirements.

21. The system of claim 19, wherein the maintenance schedule is formulated using an input data set produced by Principal Components Analysis.

22. The system of claim 19, the system further comprising a semiconductor processing equipment controller which receives data output from the neural network and including at least one of a suggested maintenance schedule for the semiconductor processing equipment, a suggested cleaning schedule for the semiconductor processing equipment and a predicted time for a next failure of the semiconductor processing equipment.

23. The system of claim 22, wherein the equipment controller uses the data output from the neural network in at least one servo control loop.

24. The system of claim 23, wherein the at least one servo control loop provides one of Proportional-integral, Proportional-Derivative, or Proportional-Integral-Derivative control, to maintain desired values of controlled variables for the semiconductor processing equipment.

25. The system of claim 16, wherein the model obtains data indicating the semiconductor equipment process measurements used for predicting the failure of the semiconductor equipment.

26. The system of claim 25, wherein the data obtained includes at least one of a time of last cleaning of the semiconductor processing equipment, a time of last maintenance of the semiconductor processing equipment, a time of a last failure of the semiconductor equipment and a preventative maintenance schedule for the semiconductor processing equipment.

27. The system of claim 16, wherein the processing measurements include at least one of chamber temperature, gas mixture or applied radio frequency power.

28. The system of claim 16, wherein the model is multivariate.

29. The system of claim 16, wherein the model is dynamic and the model is updated as processing measurement data is produced by the semiconductor processing equipment.

30. The system of claim 16, wherein the model is formulated using offline simulations to completely model the semiconductor processing equipment.

31. The system of claim 16, further comprising an icon driven user interface as a front end of the system that allows a user to retrieve data and take process measurements from the semiconductor processing equipment.

32. A method for determining a future state of semiconductor processing equipment, the method comprising:

predicting a maintenance schedule of the semiconductor processing equipment using a model determined from past performance based on semiconductor processing equipment processing measurements.

33. A system for determining a future state of semiconductor processing equipment, the system comprising:

a model determined from past performance of the semiconductor processing equipment and configured to output data indicating a maintenance schedule of the semiconductor processing equipment based on semiconductor processing equipment processing measurements.