Evaluation and ranking of manufacturing line non-numeric information

• US 5,479,361 A
• Filed: 03/24/1994
• Issued: 12/26/1995
• Est. Priority Date: 02/27/1992
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Levanoni, Menachem, Kurtzberg, Jerome M.
• Primary Examiner(s): Voeltz, Emanuel T.
• Assistant Examiner(s): KEMPER, MELANIE A
• Application Number:
  US08/217,272
• Time in Patent Office: 642 Days
• Field of Search: 364/552, 364/551.01, 364/554, 364/582, 364/581, 364/468, 364/469, 364/470-473, 364/474.02, 364/474.16, 364/474.22, 364/474.15
• US Class Current: 702/84
• CPC Class Codes: G05B 19/41875 characterised by quality surveillance of production
• CPC Class Codes: G05B 2219/32187 Correlation between controlling parameters for influence on quality parameters
• CPC Class Codes: Y02P 90/02 Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]

Abstract

An apparatus and method for quantitatively ranking the performance of each attribute contributing to a manufacturing process. Various signals representing the measure (i.e., quality or yield) of a manufacturing run and representing the attributes that contributed to each manufacturing run are recorded. An iterative process is commenced whereby numeric weights are assigned to each attribute. In a first iterative step, the weight of each attribute is determined to be the weighted average of the measure of each manufacturing run to which that attribute contributed. In subsequent iteration steps, a refined weight for each attribute is determined by computing the ratio between the normalized measure and the product of all other weights associated with that manufacturing run. The iterations are continued until the weights determined for each attribute converge or become self consistent. In this manner, the performance of non-numeric or attribute parameters such as operators and tools can be measured and defined in terms of a numeric quantity.

Claims

1. A method for optimizing a manufacturing line the manufacturing line having a plurality of non-numeric attributes, comprising the steps of:

(a) operating the manufacturing line over a plurality of manufacturing runs to produce a product, wherein each of said manufacturing runs is affected by a plurality of non-numeric attributes;

(b) recording a measure signal for each of the plurality of manufacturing runs, wherein each said measure signal indicates an output measure of the manufacturing line for its associated manufacturing run;

(c) recording attribute ID signals associated with each measured manufacturing run, wherein said attribute ID signals identify the non-numeric attributes that contributed to each manufacturing run;

(d) normalizing said recorded measure signals for each of said plurality of manufacturing runs;

(e) deconvoluting said normalized recorded measure signals and said attribute ID signals to generate a quantitative numeric measure signal for the non-numeric attributes, using a convergent iterative process; and

(f) adjusting the non-numeric attributes based on said quantitative numeric measure signal to optimize the manufacturing line, wherein production of said product is optimized as a result of said adjusting.

2. The method of claim 1, further comprising tile step of storing the results of said convergent iterative process.

3. The method of claim 1, wherein said step (c) comprises the steps of subtracting a value ε

from said recorded measure signal and dividing this result by the maximum said recorded measure signal to normalize said recorded measure signal of each manufacturing run to a value between ε and

one.

4. The method of claim 1, further comprising a step of editing manufacturing runs containing anomalous data points.

5. The method of claim 1, wherein said step (e) further comprises the steps of:

i selecting manufacturing runs to which a selected one of the plurality of non-numeric attributes contributed;

ii ascribing said measure signal for each selected run to said selected non-numeric attribute;

iii determining a weighted average of all said ascribed measure signals to generate said quantitative numeric measure signal for said selected non-numeric attribute; and

iv repeating said steps i-iii for each of said plurality of non-numeric attributes to generate said quantitative numeric measure signal for each of said plurality of non-numeric attributes.

6. The method of claim 5, further comprising the steps of:

v. computing a refined quantitative numeric measure signal for a selected one of the plurality of non-numeric attributes as a ratio between said normalized measure signal and a product of said quantitative numeric measure signals of all other non-numeric attributes; and

vi. repeating said steps i-iii for each of said plurality of non-numeric attributes to generate said refilled quantitative numeric measure signal for each of said plurality of non-numeric attributes.

7. The method of claim 6, further comprising the step of repeating said steps v--vi until said refined quantitative numeric measure signals converge.

8. A system for quantitatively determining the performance of a plurality of non-numeric attributes of a manufacturing line comprising:

(a) a manufacturing line configured to receive input materials, to process said input materials to produce a product, and to produce a measure signal that provides a quantitative measure of each manufacturing run;

(b) attributes configured to operate on said input materials to perform said material processing;

(c) means for generating attribute ID signals to identify the attributes; and

(d) a processor configured to receive said attribute ID signals and said measure signal, to normalize said recorded measure signal, and to deconvolute said normalized recorded measure signals and said attribute ID signals using a convergent iterative process to generate a quantitative numeric measure for said attributes.

9. The system of claim 8, further comprising a database for storing information from at least one of said attribute ID signals, said measure signal, said recorded measure signal, and said quantitative numeric measure.

10. The system of claim 8, further comprising means for providing additional attribute ID signals to said processor.

11. The system of claim 8, wherein said processor comprises:

i means for selecting at least one of a plurality of manufacturing runs to which a selected one of the plurality of non-numeric attributes contributed;

ii means for ascribing said measure signal for each selected run to said selected non-numeric attribute; and

iii means for determining a weighted average of all said ascribed measure signals to generate said quantitative numeric measure signal for said selected non-numeric attribute.

12. The method of claim 11, wherein said processor further comprises means for computing a refined quantitative numeric measure signal for a selected one of the plurality of non-numeric attributes as a ratio between said normalized measure signal and a product of said quantitative numeric measure signals of all other non-numeric attributes.

13. A method for quantitatively determining the performance of a plurality of non-numeric attributes of a manufacturing line, comprising the steps of:

recording a measure signal for each of a plurality of manufacturing runs on the manufacturing line, wherein each said measure signal comprises data indicating an output measure of the manufacturing line and said data are obtained by operating the manufacturing line to produce a product;

(b) recording attribute ID signals associated with each of said plurality of manufacturing runs, wherein each of said attribute ID signals identifies a non-numeric. attribute that contributed to the manufacturing run;

(c) normalizing said recorded measure signal for each of said plurality of manufacturing runs;

(d) using a convergent iterative process to deconvolute said normalized recorded measure signals and said attribute ID signals to generate a quantitative numeric measure signal for the non-numeric attributes; and

(e) adjusting the non-numeric attributes based on said quantitative numeric measure signal to optimize the manufacturing line, wherein production of said product is optimized as a result of said adjusting.

14. A system for quantitatively determining the performance of a plurality of non-numeric attributes of a manufacturing line, comprising:

(a) means for recording a measure signal for each of a plurality of manufacturing runs on the manufacturing line, wherein each said measure signal comprises data indicating an output measure of the manufacturing line and said data are obtained by operating the manufacturing line to produce a product;

(b) means for recording attribute ID signals associated with each of said plurality of manufacturing runs, wherein each of said attribute ID signals identifies a non-numeric attribute that contributed to the manufacturing run;

(c) means for normalizing said recorded measure signal for each of said plurality of manufacturing runs; and

(d) means for deconvoluting said normalized recorded measure signals and attribute ID signals using a convergent iterative process to generate a quantitative numeric measure signal for the non-numeric attributes.

15. The system of claim 14 wherein said means for deconvoluting comprises:

i. means for selecting manufacturing runs to which a selected one of the plurality of non-numeric attributes contributed;

ii. means for ascribing said recorded measure signal for each selected run to said selected non-numeric attribute; and

iii. means for determining a weighted average of all said ascribed measure signals to generate said quantitative numeric measure signal for said selected non-numeric attribute.

16. The system of claim 15, wherein said means for deconvoluting further comprises means for computing a refined quantitative numeric measure signal for a selected one of the plurality of non-numeric attributes as a ratio between said normalized measure signal and a product of said quantitative numeric measure signals of all other non-numeric attributes.

17. The system of claim 14, wherein said means for normalizing comprises means for editing manufacturing runs containing anomalous data points.