Test time reduction for manufacturing processes by substituting a test parameter

Patent metadata

Field	Value
Publication number	US-11275362-B2
Application number	US-201916433461-A
Country	US
Kind code	B2
Filing date	Jun 6, 2019
Priority date	Jun 6, 2019
Publication date	Mar 15, 2022
Grant date	Mar 15, 2022

Abstract

Methods and systems of identifying a time reduction in a manufacturing time associated with a plurality of products. One system includes an electronic processor configured to receive training data. The electronic processor is also configured to determine a first set of testing parameters from the plurality of testing parameters to remove for the assembly line based on the training data and determine a second set of testing parameters to keep by removing the first set of testing parameters from the plurality of testing parameters. The electronic processor is also configured to determine a predictive model to replace the first set of testing parameters based on the training data associated with the second set of testing parameters, and automatically update a testing process for the assembly line to turn off the first set of testing parameters and use the predictive model in place of the first set of testing parameters.

First claim

What is claimed is:

1. A system of identifying a time reduction in a manufacturing time associated with a plurality of products produced via an assembly line, the system comprising:
an electronic processor configured to
receive training data associated with a plurality of tests for the assembly line, the training data including measurements and test data associated with the plurality of tests as applied to one or more products produced via the assembly line,
determine a first set of tests from the plurality of tests to remove for the assembly line based on the training data,
determine a second set of tests to keep by removing the first set of tests from the plurality of tests,
generate a predictive model using a supervised machine learning classification method to replace the first set of tests, the predictive model trained based on the training data associated with the second set of tests, and
automatically update a testing process for the assembly line to turn off the first set of tests and use the predictive model in place of the first set of tests.

2. The system of claim 1 , wherein the electronic processor is configured to determine the first set of tests to remove for the assembly line using at least one selected from a group consisting of a principle component analysis and a regression model with a sparsity constraint.

3. The system of claim 1 , wherein the electronic processor is configured to generate the predictive model using at least one selected from a group consisting of a support vector machine, a classification and regression tree, and a boosted decision tree.

4. The system of claim 1 , wherein the electronic processor is configured to generate the predictive model by benchmarking combinations of candidate feature subsets and candidate non-linear models.

5. The system of claim 1 , wherein the electronic processor is configured to generate the predictive model by developing a customized cost-sensitive version of group least absolute shrinkage and selection operator (LASSO).

6. The system of claim 1 , wherein the electronic processor is configured to generate an additional predictive model to replace a third set of tests, wherein the third set of tests is a set of intermediate tests associated with the assembly line.

7. The system of claim 6 , wherein the additional predictive model predicts a result of an intermediate test of the assembly line and wherein the predictive model predicts a result of an end-of-the-line test of the assembly line.

8. The system of claim 1 , wherein the electronic processor is configured to substitute the first set of tests with the predictive model when a predetermined condition is satisfied.

9. The system of claim 8 , wherein the electronic processor is configured to substitute the first set of tests with the predictive model based on a confidence of the predictive model.

10. A method of identifying a time reduction in manufacturing time of a plurality of products associated with an assembly line, the method comprising:
receiving, with an electronic processor, training data associated with a plurality of tests for the assembly line, the training data including measurements and test data associated with the plurality of tests as applied to one or more products produced via the assembly line;
determining, with the electronic processor, a first set of tests from the plurality of tests to remove for the assembly line based on the training data,
determining, with the electronic processor, a second set of tests from the plurality of tests to keep by removing the first set of tests from the plurality of tests,
generating, with the electronic processor, a predictive model using a supervised machine learning classification method to replace the first set of tests, the predictive model trained based on the training data associated with the second set of tests, and
automatically updating a testing process for the assembly line to turn off the first set of tests and use the predictive model in place of the first set of tests.

11. The method of claim 10 , further comprising:
generating an additional predictive model to replace a third set of tests, wherein the third set of tests is a set of intermediate tests associated with the assembly line.

12. The method of claim 11 , wherein generating the additional predictive model includes generating an additional predictive model predicting a result of an intermediate test of the assembly line and wherein the predictive model predicts a result of an end-of-the-line test of the assembly line.

13. The method of claim 10 , further comprising:
substituting the first set of tests with the predictive model when a predetermined condition is satisfied.

14. The method of claim 13 , wherein substituting the first set of tests with the predictive model when the predetermined condition is satisfied includes substituting the first set of tests with the predictive model based on a confidence of the predictive model.

15. The method of claim 10 , wherein determining the first set of tests to remove for the assembly line includes determining the first set of tests to remove for the assembly line using at least one selected from a group consisting of a principle component analysis and a regression model with a sparsity constraint.

16. A non-transitory computer readable medium including instructions that, when executed by an electronic processor, causes the electronic processor to execute a set of functions, the set of functions comprising:
receiving training data associated with a plurality of tests for an assembly line, the training data including measurements and test data associated with the plurality of tests as applied to one or more products produced via the assembly line;
determining a first set of tests from the plurality of tests to remove for the assembly line based on the training data;
determining a second set of tests from the plurality of tests to keep by removing the first set of tests from the plurality of tests;
generating a predictive model using a supervised machine learning classification method to replace the first set of tests, the predictive model trained based on the training data associated with the second set of tests; and
automatically updating a testing process for the assembly line to turn off the first set of tests and use the predictive model in place of the first set of tests.

17. The computer readable medium of claim 16 , wherein the set of functions further comprises:
generating an additional predictive model to replace a third set of tests using data mining, wherein the third set of tests is a set of intermediate tests associated with the assembly line.

18. The computer readable medium of claim 17 , wherein generating the additional predictive model includes generating an additional predictive model predicting a result of an intermediate test of the assembly line and wherein the predictive model predicts a result of an end-of-the-line test of the assembly line.

19. The computer readable medium of claim 16 , wherein the set of functions further comprises:
substituting the first set of tests with the predictive model when a predetermined condition is satisfied.

20. The computer readable medium of claim 19 , wherein substituting the first set of tests with the predictive model includes substituting the first set of tests with the predictive model based on a confidence of the predictive model.

Assignees

• Bosch Gmbh Robert

Inventors

• Ghosh Rumi
• Srinivasan Soundararajan
• Chen Ruobing
• Kang Shan
• Naumann Marc
• Tandarpally Mahesh Goud

Classifications

• G06N5/01

  Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound · CPC title

• G06F18/24

  Classification techniques · CPC title

• G06F18/214

  Generating training patterns; Bootstrap methods, e.g. bagging or boosting · CPC title

• G06N20/20

  Ensemble learning · CPC title

• G06N20/10

  using kernel methods, e.g. support vector machines [SVM] · CPC title