Adaptive Test Program Generation Using Neural Network Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing memory-testing procedures require significant human resources and costly machines, and they fail to address quality loss and low yield due to variations in manufacturing processes, as they cannot efficiently adjust test programs to achieve optimal test coverage within reasonable time.
Innovation Solution
A testing system that includes test equipment and a testing-control apparatus, which executes a test program and uses a neural network to analyze test results to generate an adaptive test program for the next iteration, optimizing the combination and order of test items to reduce testing time and increase coverage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual test program design and adjustment is used, then test coverage can be achieved within reasonable test time, but significant human resources are required and it cannot solve quality loss due to manufacturing variations
Solution Approach 1:
The system enables self-service by automatically generating and optimizing test programs through the neural network without requiring manual intervention. The testing-control apparatus autonomously analyzes test results and adjusts test parameters, allowing the system to serve itself in improving test coverage while eliminating dependency on human operators.
Solution Approach 2:
The system implements feedback by continuously analyzing test results through the neural network and using this information to dynamically adjust and optimize subsequent test programs. This closed-loop feedback mechanism enables automatic adaptation to manufacturing variations, improving reliability without manual intervention while maintaining high test coverage.
2Reliability
If traditional testing procedures are used, then testing can be performed, but it requires costly machines and cannot address quality loss due to manufacturing process variations
Solution Approach 1:
The system applies dynamics by making the test program adaptable and changeable based on real-time test results and manufacturing variations. The neural network dynamically adjusts test parameters and sequences, enabling the system to respond to changing conditions rather than following fixed procedures, thus improving quality consistency across different manufacturing variations.
Solution Approach 2:
The system implements parameter changes by modifying test program parameters based on neural network analysis of test results. This includes adjusting test conditions, sequences, and configurations to optimize for specific manufacturing variations, thereby improving reliability and quality consistency without requiring expensive specialized equipment for each variation.
3Reliability
If comprehensive test items are executed to ensure quality, then test coverage increases, but testing time increases significantly
Solution Approach 1:
The system applies partial action by selectively executing only the necessary test items based on neural network analysis of previous test results. Instead of running all possible test items, the system identifies and executes only those tests that provide the most value for achieving quality goals, thereby reducing testing time while maintaining adequate coverage through intelligent selection rather than exhaustive testing.
Data Source
AI summary
A testing system is provided. The testing system includes: test equipment and a testing-control apparatus. The test equipment is configured to perform tests on a device under test. The testing-control apparatus is configured to execute a test program to control the test equipment to perform a plurality of first test items in the test program on the device under test. The testing-control apparatus retrieves a test result of each of the first test items from the test equipment, and executes a test-program neural network to analyze the test result of each of the first test items to generate the test program for a next test iteration.