Computer-implemented system and method of performing a design of experiment, generating a product thereof, generating designed experiment options, and a graphical user interface

• US 9,965,721 B2
• Filed: 09/16/2014
• Issued: 05/08/2018
• Est. Priority Date: 09/16/2013
• Status: Active Grant

• Current Assignee: Moresteam Development LLC
• Original Assignee: Moresteam Development LLC
• Inventors: Maaseidvaag, Lars, Skrivanek, Smita Kulkarni, Gao, Alan
• Primary Examiner(s): Hill, Stanley K.
• Application Number:
  US14/488,060
• Publication Number:
  US 20150081596A1
• Time in Patent Office: 1,330 Days
• Field of Search: None
• CPC Class Codes: G06N 5/04 Inference or reasoning models

Abstract

Disclosed is a computer-implemented method of generating a plurality of selectable design experiments in a design of experiments (DOE) process for analyzing at least one data set from a process to determine a relationship of a plurality of process factors of interest to a process output of interest. The method entails receiving as user input into a computing apparatus, a base number of trials and a number of center point trials. From the computing apparatus, a graphical user interface is generated on a display connected with the computing apparatus. This includes presenting a display containing a plurality of experiment each defined, at least partially, by a combination of number of replicates (replicates count) and effect size. The display further includes a predicted power value for each combination of replicates count and effect size.

Claims

1. A computer-implemented method of generating a plurality of selectable design experiments in the conduct of a design of experiments (DOE) process for analyzing a data set from a process to determine an effects relationship between a plurality of process factors and a process output, said method comprising:

receiving as user input into a computing apparatus, a base number of trials and a number of center point trials; and

generating a graphical user interface on a display connected with the computing apparatus, including presenting, at once, on the graphical user interface, a plurality of mutually exclusive designed experiments each defined, at least partially, by a combination of number of replicates count and effect size, and a predicted power value identified to each combination of replicates count and effect size, wherein each predicted power value is a value in a range from 0% to 100%, wherein generation of the graphical user interface is responsive to the user input of the base number of trials and the number of center point trials, and wherein presenting the designed experiments includes presenting the number of replicates count and effect size, and the predicted power value for each designed experiment; and

generating a table identifying a power value for each combination of replicates count and effect size and wherein the table contains a plurality of cells, each containing a power value, each cell being associated uniquely with one of said combinations of replicates count and effect size, wherein presenting the plurality of designed experiment includes presenting the table.

2. The method of claim 1, further comprising:

generating a graphical user interface display to elicit said user input of the base number of trials and the number of center point trials from a user and wherein the power values are contained in cells forming a matrix, and wherein each cell is engageable by a user to select a designed experiment option identified to the combination of replicates count and effect size associated with said engaged cell.

3. The method of claim 1, wherein the computing apparatus includes a non-transitory computer-accessible storage medium having program instructions stored thereon, wherein the program instructions are computer-executable to implement generate and present the graphical user interface and the table for presentation on the display.

4. The method of claim 1, further comprising:

upon selection of a designed experiment option by a user selecting a cell that identifies a power value, generating a subsequent graphical user interface presenting information on the selected designed experiment option.

5. The method of claim 4, wherein cells containing power values above a threshold value are displayed with a first color and cells displayed with a second color contain power values less than the threshold value.

6. The method of claim 1, wherein presenting designed experiments includes displaying a cost of trials value for each said combination of replicates count and effect size.

7. The method of claim 1, further comprising:

prior to presenting designed experiment options, generating a graphical user interface to receive factor information including a factor designation, and high level and low level values for each factor designation.

8. The method of claim 7, further comprising:

prior to presenting designed experiments, generating a graphical user interface to receive, as user input, a base number of trials.

9. The method of claim 8, wherein generating a user interface to receive a base number of trials, includes eliciting user election of either a full experiment or a fractional experiment.

10. The method of claim 8, further comprising:

prior to presenting designed experiments, generating a graphical user interface to receive, as input to a determination of power values, a numerical value for number of center points, high level values and low level values for each of a plurality of selected factors.

11. The method of claim 10, further comprising:

prior to presenting designed experiments, generating a graphical user interface to receive, as input to a determination of power values, a numerical value for target minimum effect size.

12. The method of claim 10, further comprising:

prior to presenting designed experiments, generating a graphical user interface to receive, as input to a determination of power values, a numerical value for a standard deviation of the process.

13. The method of claim 8, further comprising:

prior to presenting designed experiments, generating one or more user interfaces to receive, as input to a determination of power values, a numerical value for a number of center points, a numerical value for target minimum effect size, and base number of trials.

14. The method of claim 1, wherein presenting designed experiments includes generating a matrix defined by replicates count and effect size, whereby the matrix contains, as elements, predicted power values for each combination of replicates count and effect size.

15. The method of claim 14, wherein presenting designed experiments includes generating a cost of trials column aligned with a column of the replicates count.

16. A non-transitory graphical user interface generated by a computing apparatus, said graphical user interface comprising:

a section presenting experiment options including a first array of numerical values of replicates count in a designed experiment option and a second array of numerical values of effect size in a designed experiment option, wherein the first array and second array are positioned in alignment such that each of the replicates counts in the first array uniquely align with one of the effect size values in the second array, and wherein a designed experiment power value is determined and presented in unique association with a combination of one of said replicates count and the effect sizes aligned therewith; and

wherein the first array, the second array, and the power values are presented in a matrix table having a dimension equal to the number of replicate values by the number of effect size values, and containing, as elements, power values for each said combination of replicates count and effect size and wherein the power values are contained in cells forming the matrix, and wherein each cell is engageable by a user to select a designed experiment option identified to the combination of replicates count and effect size associated with said engaged cell and generate a user interface associated with the engaged cell.

17. A non-transitory computer-accessible storage medium storing program instructions computer-executable to implement one or more graphical user interfaces for presentation of design of experiments options on a computer display, wherein the experiments are for analyzing a data set from a process to determine an effects relationship between a plurality of process factors and a process output, wherein:

generation of a graphical user interface presenting designed experiment options is responsive to user select input of base number of trials, effect size, and number of center point trials; and

the graphical user interface presenting designed experiment options includes a first array of numerical values of replicates count in a designed experiment option and a second array of numerical values of effect size in a designed experiment option; and

wherein the first array and second array are positioned in alignment such that each of the replicates counts in the first array uniquely align with one of the effect size values in the second array, and wherein a designed experiment power value is determined and presented in unique association with a combination of one of said replicates count and the effect sizes aligned therewith.

18. The computer-accessible medium of claim 17, wherein the graphical user interface presenting designed experiment options further includes an array of cost of trial values, each of the cost values being associated with one of said combinations of replicates count and effect size.

19. The computer-accessible medium of claim 17, wherein the power values are color coded by strength of power value.

20. The computer-accessible medium of claim 17, wherein the first array, the second array, and the power values are presented in a matrix table having a dimension equal to the number of replicate values by the number of effect size values, and containing, as elements, power values for each said combination of replicates count and effect size.

21. The computer-accessible medium of claim 17, wherein one or more user interfaces preceding the user interface presenting designed experiment options includes a first user interface configured to receive, as user input, base number of trials information, a second user interface configured to receive, as user input, factor and level information, and a third user interface configured to receive, as input, target effect size range;

and

wherein said power values are calculated based, at least partly, on values for base number of trial and target size effect as inputted on said first, second, and third user interfaces.

22. A computer-implemented method of conducting a design of experiments (DOE) for analyzing a data set from a process to determine an effects relationship between a plurality of process factors and a process output, said method comprising:

entering into a computing apparatus, a base number of trials, a number of center point trials, and a target effect size;

prompting the computing apparatus to generate on a display connected with the computing apparatus, a graphical user interface, including presenting a display containing a plurality of user-responsive cells each presenting a designed experiment option defined, at least partially, by a combination of replicates count and effect size, the user interface further presenting a predicted power value for each combination of replicates count and effect size;

wherein presenting the plurality of designed experiments further includes generating a table identifying a power value for each combination of replicates count and effect size;

wherein the table contains a plurality of cells, each containing a power value, each cell being associated uniquely with one of said combinations of replicates count and effect size; and

wherein, upon selection of a designed experiment option by a user selecting a cell that contains a power value, generating a subsequent graphical user interface presenting information on the selected designed experiment option.

23. The method of claim 22, wherein presenting designed experiment options further comprises generating a table displaying the power values, wherein each power value is identified uniquely by a combination of replicates count and effect size.

24. The method of claim 23, wherein the table contains a plurality of cells, each cell containing a power value from 0% to 100% identified to one of said combination of replicates count and effect size.

25. The method of claim 24, further comprising:

selecting an experiment option by selecting a cell containing a power value, thereby causing display of a subsequent graphical user interface presenting information on the selected designed experiment option.

26. The method of claim 25, wherein cells with power values above a threshold value are displayed with a first color and cells displayed with a second color contain power values less than the threshold value.

27. The method of claim 24, wherein presenting designed experiment options includes generating a matrix defined by replicates count and effect size, whereby the matrix contains, as elements, predicted power values for each combination of replicates count and effect size.

28. The method of claim 27, further comprising modifying presentation of the matrix on the graphical user interface by engaging a vertical axis defined by replicates count and causing the vertical axis to extend, thereby adding a plurality of rows to the matrix and increasing a total number of replicates counts and thereby increasing the number of power values as elements of the matrix, and increasing the number of designed experiment options presented.

29. The method of claim 22, further comprising generating a modified table in response to changes to the number of replicates and effect size.

30. The method of claim 22, further comprising generating a modified table in response to changes to the number of replicates and effect size to attain at least one design experiment having a predicted power value of 100%.

31. The method of claim 30, wherein said generating step is repeated multiple times until a table having a cell with a predicted power value of 100% is generated.