Systems and methods for reducing manufacturing failure rates

Patent metadata

Field	Value
Publication number	US-11954607-B2
Application number	US-202217992769-A
Country	US
Kind code	B2
Filing date	Nov 22, 2022
Priority date	May 9, 2017
Publication date	Apr 9, 2024
Grant date	Apr 9, 2024

Abstract

Systems and methods are provided for reducing failure rates of a manufactured products. Manufactured products may be clustered together according to similarities in their production data. Manufactured product clusters may be analyzed to determine mechanisms for failure rate reduction, including adjustments to test quality parameters, product formulas, and product processes. Recommended product adjustments may be provided.

First claim

What is claimed is:

1. A system for reducing failure rates of a manufactured product comprising:
one or more processors; and
memory storing instructions that, when executed by the one or more processors, cause the system to perform:
obtaining, from a first database, one or more product data sets corresponding to a product, each of the one or more product data sets comprising one or more constituents;
obtaining, from a second database, time series sensor data acquired from sensors during production of the product;
clustering the one or more product data sets based on the one or more constituents;
predicting one or more results of failures of the product based on the clustering;
acquiring one or more actual results of failures based on information from the second database; and
based on a comparison between the actual results of the failures and the predicted results of the failures, generating a model that correlates product constituents and amounts of each of the product constituents with a test result for the product and outputting an adjustment to the product formula to reduce a failure rate of the product formula;
receiving an input of a new formula according to the outputted adjustment to the product formula; and
evaluating the new formula based on the generated model; and
recommending any alternations to the new formula based on the evaluation of the new formula.

2. The system of claim 1 , wherein the time series sensor data comprises a temperature time series data, a pressure time series sensor data, and a mixing speed time series data.

3. The system of claim 1 , wherein the outputting an adjustment to the product formula comprises:
in response to determining that the actual results of failures and the predicted results of failures correspond to a same relative position with respect to a quality parameter range, determining to output the adjustment to the product formula.

4. The system of claim 3 , wherein the outputting an adjustment to the product formula comprises:
generating an adjusted product formula by adjusting an amount of a constituent in the product formula.

5. The system of claim 3 , wherein the predicted results of failures corresponding to the same relative position with respect to a quality parameter range as the actual results of failures indicates that:
the predicted results of failures and the actual results of failures are attributed to results above the quality parameter range, or
the predicted results of failures and the actual results of failures are attributed to results below the quality parameter range.

6. The system of claim 3 , wherein the quality parameter range indicates a lower bound and an upper bound for a passing test result of a parameter.

7. The system of claim 1 , wherein the instructions further cause the system to perform:
normalizing the one or more actual results of failures based on same time windows as successful product runs.

8. A method comprising:
obtaining, from a first database, one or more product data sets corresponding to a product, each of the one or more product data sets comprising one or more constituents;
obtaining, from a second database, time series sensor data acquired from sensors during production of the product;
clustering the one or more product data sets based on the one or more constituents;
predicting one or more results of failures of the product based on the clustering, wherein the clustering or the predicting is performed by a machine learning technique;
acquiring one or more actual results of failures based on information from the second database; and
based on a comparison between the actual results of the failures and the predicted results of the failures, generating a model that correlates product constituents and amounts of each of the product constituents with a test result for the product and outputting an adjustment to the product formula to reduce a failure rate of the product formula;
receiving an input of a new formula according to the outputted adjustment to the product formula; and
evaluating the new formula based on the generated model; and
recommending any alternations to the new formula based on the evaluation of the new formula.

9. The method of claim 8 , wherein the time series sensor data comprises a temperature time series data, a pressure time series sensor data, and a mixing speed time series data.

10. The method of claim 8 , wherein the outputting an adjustment to the product formula comprises:
in response to determining that the actual results of failures and the predicted results of failures correspond to a same relative position with respect to a quality parameter range, determining to output the adjustment to the product formula.

11. The method of claim 10 , wherein the outputting an adjustment to the product formula comprises:
generating an adjusted product formula by adjusting an amount of a constituent in the product formula.

12. The method of claim 10 , wherein the predicted results of failures corresponding to the same relative position with respect to a quality parameter range as the actual results of failures indicates that:
the predicted results of failures and the actual results of failures are attributed to results above the quality parameter range, or
the predicted results of failures and the actual results of failures are attributed to results below the quality parameter range.

13. The method of claim 10 , wherein the quality parameter range indicates a lower bound and an upper bound for a passing test result of a parameter.

14. A non-transitory computer readable medium comprising instructions that, when executed, cause one or more processors to perform:
obtaining, from a first database, one or more product data sets corresponding to a product, each of the one or more product data sets comprising one or more constituents;
obtaining, from a second database, time series sensor data acquired from sensors during production of the product;
clustering the one or more product data sets based on the one or more constituents;
predicting one or more results of failures of the product based on the clustering;
acquiring one or more actual results of failures based on information from the second database; and
based on a comparison between the actual results of the failures and the predicted results of the failures, generating a model that correlates product constituents and amounts of each of the product constituents with a test result for the product and outputting an adjustment to the product formula to reduce a failure rate of the product formula;
receiving an input of a new formula according to the outputted adjustment to the product formula; and
evaluating the new formula based on the generated model; and
recommending any alternations to the new formula based on the evaluation of the new formula.

15. The non-transitory computer readable medium of claim 14 , wherein the time series sensor data comprises a temperature time series data, a pressure time series sensor data, and a mixing speed time series data.

16. The non-transitory computer readable medium of claim 14 , wherein the outputting an adjustment to the product formula comprises:
in response to determining that the actual results of failures and the predicted results of failures correspond to a same relative position with respect to a quality parameter range, determining to output the adjustment to the product formula.

Assignees

• Palantir Technologies Inc

Inventors

• Seaton William
• Wiltsche Clemens
• Novick Myles
• Patel Rootul

Classifications

• G06N5/022 Primary

  Knowledge engineering; Knowledge acquisition · CPC title

• G06Q10/06 Primary

  Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling · CPC title