# US 12,007,428 B2 · Systems and methods for multidimensional dynamic part average testing > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,007,428 B2 - Document ID: US12007428B2 - Status: Granted - Kind: B2 - Application number: 17/497,518 - Filed: 2021-10-08 - Published: 2023-04-13 - Granted: 2024-06-11 - Earliest claimed: 2021-10-08 - Assignees: Advantest Corporation - Applicants: Advantest Corporation - Inventors: Kenneth Butler; Ira Leventhal; Constantinos Xanthopoulos; Alan Hart; Brian Buras; Keith Schaub - Web page: https://patentgrep.com/patent/US12007428B2 ## CPC - G01R 31/26: Testing of individual semiconductor devices (testing of photovoltaic devices H02S50/10; testing or measuring during manufacture or treatment {H10P74/00}) - G01R 31/2601: {Apparatus or methods therefor (G01R31/2607, G01R31/2642 take precedence)} - G01R 31/01: Subjecting similar articles in turn to test, e.g. "go/no-go" tests in mass production; Testing objects at points as they pass through a testing station (testing of cables continuously passing the testing apparatus G01R31/59; testing dielectric strength or breakdown voltage G01R31/12) - G01R 31/31718: {Logistic aspects, e.g. binning, selection, sorting of devices under test, tester/handler interaction networks, Test management software, e.g. software for test statistics or test evaluation, yield analysis (mechanical aspects G01R31/2808, G01R31/2851)} - G01R: MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES (indicating correct tuning of resonant circuits H03J3/12) ## Abstract Embodiments of the present invention provide systems and methods for multidimensional parts average testing for testing devices and analyzing testing results to detect outliers according to embodiments of the present invention. The testing can include calculating multivariate (e.g., bivariate) statistics using delta measurements of like devices, a ratio of measurements, or principal component analysis that identifies eigenvectors and eigenvalues to define meta parameters, for example. Raw test result data can be converted to residual space and robust regression can be performed to prevent outlier results from influencing regression, thereby reducing overkill advantageously. ## Claims ### Claim 1 A method of dynamic parts average testing, the method comprising: determining testing limits based on historic testing result data; testing a plurality of portions of a material according to the testing limits to obtain test results; computing multivariate statistics using the test results; computing mean or median values of the multivariate statistics; computing differences between the multivariate statistics and the mean or median values; updating the testing limits based on the mean or median values to produce updated testing limits; identifying a trend indicating an out of control material drift of the material based on the mean or median values and the updated testing limits; and halting testing of the plurality of portions of the material responsive to the identifying the trend. ### Claim 2 (depends on 1) The method as described in claim 1, further comprising determining significant measurements for testing by performing at least one of: principal component analysis (PCA); independent component analysis (ICA); auto encoding; or machine learning. ### Claim 3 (depends on 1) The method as described in claim 1, wherein the computing multivariate statistics using the test results comprises at least one of: forming pairs; forming ratios; and forming deltas. ### Claim 4 (depends on 1) The method as described in claim 1, wherein the computing multivariate statistics using the test results comprises clustering the test results according to a result type. ### Claim 5 (depends on 1) The method as described in claim 1, further comprising converting the mean or median values to residual space using a non-linear, non-monotonic transformation to amplify outlier results. ### Claim 6 (depends on 1) The method as described in claim 1, further comprising removing outlier results before the updating the testing limits based on the mean or median values. ### Claim 7 (depends on 1) The method as described in claim 1, further comprising performing reweighted least square regression before the updating the testing limits based on the mean or median values. ### Claim 8 An apparatus for performing dynamic parts average testing, the apparatus comprising: a processor; and a memory in communication with the processor for storing test data and instructions, wherein the processor executes instructions to perform a method of multivariate parts average testing, the method comprising: determining testing limits based on historic testing result data; testing a plurality of portions of a material according to the testing limits to obtain test results; computing multivariate statistics using the test results; computing mean or median values of the multivariate statistics; computing differences between the multivariate statistics and the mean or median values; updating the testing limits based on the mean or median values to produce updated testing limits; identifying a trend indicating an out of control material drift of the material based on the mean or median values and the updated testing limits; and halting testing of the plurality of portions of the material responsive to the identifying the trend. ### Claim 9 (depends on 8) The apparatus as described in claim 8, wherein the method further comprises determining significant measurements for testing by performing at least one of: principal component analysis (PCA); auto encoding; or machine learning. ### Claim 10 (depends on 8) The apparatus as described in claim 8, wherein the computing multivariate statistics using the test results comprises at least one of: forming bivariate pairs; forming bivariate ratios; and forming bivariate deltas. ### Claim 11 (depends on 8) The apparatus as described in claim 8, wherein the computing multivariate statistics using the test results comprises clustering the test results according to a result type. ### Claim 12 (depends on 8) The apparatus as described in claim 8, wherein the method further comprises converting the mean or median values to residual space using a non-linear, non-monotonic transformation to amplify outlier results. ### Claim 13 (depends on 8) The apparatus as described in claim 8, wherein the method further comprises removing outlier results before the updating the testing limits based on the mean or median values. ### Claim 14 (depends on 8) The apparatus as described in claim 8, wherein the method further comprises performing reweighted least square regression before the updating the testing limits based on the mean or median values. ### Claim 15 One or more non-transitory computer-readable medium comprising program instructions stored thereon that are executable by one or more processors to implement a method comprising: determining testing limits based on historic testing result data; testing a plurality of portions of a material according to the testing limits to obtain test results; computing multivariate statistics using the test results; computing mean or median values of the multivariate statistics; computing differences between the multivariate statistics and the mean or median values; updating the testing limits based on the mean or median values to produce updated testing limits; identifying a trend indicating an out of control material drift of the material based on the mean or median values and the updated testing limits; and halting testing of the plurality of portions of the material responsive to the identifying the trend. ### Claim 16 (depends on 15) The non-transitory computer-readable medium as described in claim 15, wherein the method further comprises determining significant measurements for testing by performing at least one of: principal component analysis (PCA); independent component analysis (ICA); auto encoding; or machine learning. ### Claim 17 (depends on 15) The non-transitory computer-readable medium as described in claim 15, wherein the computing multivariate statistics using the test results comprises at least one of: forming bivariate pairs; forming ratios; and forming deltas. ### Claim 18 (depends on 15) The non-transitory computer-readable medium as described in claim 15, wherein the computing multivariate statistics using the test results comprises clustering the test results according to a result type. ### Claim 19 (depends on 15) The non-transitory computer-readable medium as described in claim 15, wherein the method further comprises converting the mean or median values to residual space using a non-linear, non-monotonic transformation to amplify outlier results. ### Claim 20 (depends on 15) The non-transitory computer-readable medium as described in claim 15, wherein the determining testing limits based on historic testing result data comprises accessing at least one of: e-test data; data produced during a previous manufacturing step; statistically correlated data; or topologically correlated data.