Determination apparatus, test system, determination method, and computer-readable medium

1 . A testing apparatus comprising:

a main body;

a test head operatively connected to the main body configured to transmit test signals to a device under measurement; and

a determination apparatus having at least one processor configured to:

acquire test results of tests performed using the test head on a plurality of items which are performed on the device under measurement;

perform a first determination as to whether to retest the device under measurement that has failed the tests; and

if it is determined to retest the device under measurement, signal to the main body to retest the device under measurement;

wherein

the at least one processor:

performs the first determination based on reproducibility of the test results in a case where the tests have been performed on a plurality of ones of the devices under measurement multiple times;

performs the first determination, based on the reproducibility, where ones of the device under measurement that have failed a test of a corresponding item;

performs the first determination by using a learning model that has learned the reproducibility;

outputs a prediction result of the retest according to inputs of the test results of the plurality of items;

supplies, to the learning model, the test results of the acquired plurality of items of ones of the device under measurement;

executes learning processing of the learning model by using training data including at least an item ID of a failed item among the plurality of items, and a result of a retest; and

acquires the reproducibility from the prediction result of the retest which is output by the learning model according to the test results of the plurality of items being supplied to the learning model.

2 . The testing apparatus according to claim 1 , wherein the at least one processor is further configured to:

perform a second determination relating to whether to retrain the learning model by using the result of the retest and the first determination, and;

execute the learning processing of the learning model according to the second determination.

3 . The testing apparatus according to claim 1 , wherein the at least one processor is further configured to

calculate the reproducibility from the test results of the plurality of tests on ones of the device under measurement and test results of a plurality of retests.

4 . The testing apparatus according to claim 3 , wherein the at least one processor is further configured to:

perform a second determination as to whether to update the reproducibility by using a result of the retest and the first determination and

update the reproducibility according to the second determination.

5 . The testing apparatus according to claim 2 , wherein the at least one processor is further configured to:

perform a third determination as to whether to retest the device under measurement, regardless of the first determination, according to tests being performed on ones of the device under measurement included in a lot at an interval of a reference number, and

perform the second determination by using a result of the retest.

6 . The testing apparatus according to claim 4 , wherein the at least one processor is further configured to:

perform a third determination as to whether to retest the device under measurement, regardless of the first determination, according to the tests being performed on ones of the device under measurement included in a lot at an interval of a reference number, and

perform the second determination by using a result of the retest.

7 . The testing apparatus according to claim 1 , wherein the at least one processor is further configured to:

acquire an allowable lower limit value of a rate of a device under measurement that passes the test; and

calculate, based on the reproducibility of each item, among devices under measurement that have failed the test of a corresponding item, a number of at least one predicted pass of a device under measurement that is predicted to pass at least the retest of the item, wherein

as a threshold value, the first testing, in a case of integrating the number of at least one predicted pass in order according to a magnitude of corresponding reproducibility of each item, a value based on the reproducibility corresponding to the number of at least one predicted pass when or immediately before an integration result reaches a number corresponding to the allowable lower limit value.

8 . The testing apparatus according to claim 1 , wherein the at least one processor is further configured to:

acquire an allowable lower limit value of a rate of a device under measurement that passes the test; and

calculate, based on the reproducibility of each item, among devices under measurement that have failed the test of a corresponding item, a number of at least one predicted pass of a device under measurement that is predicted to pass at least the retest of the item, wherein

as a threshold value, the first determination uses, in a case of integrating the number of at least one predicted pass in order according to a magnitude of corresponding reproducibility of each item, a value based on the reproducibility corresponding to the number of at least one predicted pass when or immediately before an integration result reaches a number corresponding to the allowable lower limit value.

9 . The testing apparatus according to claim 1 , wherein the at least one processor is further configured to:

acquire an allowable lower limit value of a rate of a device under measurement that passes the test; and

calculate, based on the reproducibility of each item, among devices under measurement that have failed the test of a corresponding item, a number of at least one predicted pass of a device under measurement that is predicted to pass at least the retest of the item, wherein

as a threshold value, the first determination uses, in a case of integrating the number of at least one predicted pass in order according to a magnitude of corresponding reproducibility of each item, a value based on the reproducibility corresponding to the number of at least one predicted pass when or immediately before an integration result reaches a number corresponding to the allowable lower limit value.

10 . The testing apparatus according to claim 1 , wherein the at least one processor is further configured to:

acquire a data file including the test results, and extract the test results of each item from the data file; and

store the extracted test results in storage, wherein

the test results of the plurality of items are acquired from the storage.

11 . The testing apparatus according to claim 1 , wherein the at least one processor is further configured to:

acquire a data file including the test results, and extract the test results of each item from the data file; and

store the extracted test results in storage, wherein

the test results of the plurality of items are acquired from the storage.

12 . The testing apparatus according to claim 1 , wherein the at least one processor is further configured to:

acquire a data file including the test results, and extract the test results of each item from the data file; and

store the extracted test results in storage, wherein

the acquired test results of the plurality of items are acquired from the storage.

13 . A test system comprising:

the determination apparatus according to claim 1 ; and

a test apparatus that uses the at least one processor to perform the tests on the plurality of items on the device under measurement.

14 . A testing method comprising:

acquiring test results of tests performed using a test head configured to transmit test signals on a plurality of items which are performed on a device under measurement;

first determining of whether to retest the device under measurement that has failed the tests and;

if it is determined to retest the device under measurement, signaling to a testing apparatus to retest the device under measurement;

wherein

in the first determining, the determination is performed based on reproducibility of the test results in a case where the tests have been performed on a plurality of ones of the device under measurement multiple times;

in the first determining, the determination is performed, based on the calculated reproducibility, whether to retest ones of the device under measurement that have failed a test of a corresponding item; and

in the first determining, the determination is performed by using a learning model that has learned the reproducibility;

outputting a prediction result of the retest according to inputs of the test results of the plurality of items;

supplying, to the learning model, the test results of the acquired plurality of items of ones of the device under measurement;

executing learning processing of the learning model by using training data including at least an item ID of a failed item among the plurality of items, and a result of a retest; and

acquiring the reproducibility from the prediction result of the retest which is output by the learning model according to the test results of the plurality of items being supplied to the learning model.

15 . A non-transitory computer-readable medium having stored thereon a testing program that is executed by a computer having at least one processor configured to:

acquire test results performed using a test head configured to transmit test signals of tests on a plurality of items which are performed on a device under measurement;

perform a first determination as to whether to retest the device under measurement that has failed the tests; and

if it is determined to retest the device under measurement, signal to a testing apparatus to retest the device under measurement;

wherein

the at least one processor:

performs the first determination based on reproducibility of the test results in a case where the tests have been performed on a plurality of ones of the devices under measurement multiple times;

performs the first determination, based on the reproducibility, where ones of the device under measurement that have failed a test of a corresponding item;

performs the first determination by using a learning model that has learned the reproducibility;

outputs a prediction result of the retest according to inputs of the test results of the plurality of items;

supplies, to the learning model, the test results of the acquired plurality of items of ones of the device under measurement;

executes learning processing of the learning model by using training data including at least an item ID of a failed item among the plurality of items, and a result of a retest; and

acquires the reproducibility from the prediction result of the retest which is output by the learning model according to the test results of the plurality of items being supplied to the learning model.

16 . A testing apparatus comprising:

a main body;

a test head operatively connected to the main body configured to transmit test signals to a device under measurement; and

a determination apparatus having at least one processor configured to:

acquire test results of tests performed using the test head on a plurality of items which are performed on the device under measurement;

perform a first determination as to whether to retest the device under measurement that has failed the tests; and

if it is determined to retest the device under measurement, signal to the main body to retest the device under measurement;

wherein

the at least one processor:

performs the first determination based on reproducibility of the test results in a case where the tests have been performed on a plurality of ones of the devices under measurement multiple times;

performs the first determination, based on the reproducibility, where ones of the device under measurement have failed a test of a corresponding item;

performs the first determination by using a learning model that has learned the reproducibility;

calculates the reproducibility from the test results of the plurality of tests on ones of the device under measurement and test results of a plurality of retests;

performs a second determination as to whether to update the reproducibility by using a result of the retest and the first determination; and

updates the reproducibility according to the second determination.