Inspection Standard Setting Device for Defect Sign Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Setting an appropriate inspection standard to detect defect signs in intermediate products is challenging due to the difficulty in quantifying the causal correlation between defect signs and defect generation in subsequent processes, leading to issues with yield rate and reinspection costs.
Innovation Solution
A technique that uses a process inspection device to extract features from intermediate products and compare them with predetermined standards, employing an inspection standard setting device to calculate the separation degree between good and defective products, allowing for the selection and adjustment of inspection standards based on causal correlation, and optimizing these standards to maximize cost-effectiveness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the inspection standard is made stricter to detect more defect signs, then the detection capability is improved, but the yield rate is lowered and reinspection costs increase
Solution Approach 1:
The patent applies parameter changes by adjusting inspection standards based on calculated separation degrees. Instead of uniformly strict standards, the system dynamically modifies inspection parameters (thresholds, criteria) for specific inspection items where defect signs are strongly correlated with final defects, thereby improving detection precision without unnecessarily reducing yield rate across all items.
Solution Approach 2:
The inspection standard setting is made dynamic through automatic adjustment based on separation degree calculations. The system continuously learns from final inspection results and adapts inspection standards in real-time, transitioning from static predetermined standards to dynamic data-driven standards that optimize both detection capability and yield rate.
2Measurement precision
If the inspection standard is made stricter to detect more defect signs, then the detection capability is improved, but the reinspection costs increase
Solution Approach 1:
The system changes inspection parameters selectively based on separation degree analysis. By identifying which inspection items have strong causal correlation with final defects, the system adjusts parameters only for those specific items, avoiding unnecessary strict inspection across all items and thereby reducing reinspection costs while maintaining detection capability for critical defect signs.
Solution Approach 2:
The patent applies partial action by focusing inspection standard adjustments only on inspection items that show strong separation degree with final defects. Instead of applying strict inspection universally, the system selectively intensifies inspection only where it provides maximum value, reducing overall reinspection costs while maintaining effective defect sign detection.
3Measurement precision
If there are several inspection items with multiple standards, then the inspection coverage is improved, but it becomes difficult to specify which inspection standard was effectively adjusted
Solution Approach 1:
The system implements feedback by automatically calculating separation degrees using final inspection results and feeding this information back to adjust inspection standards. This closed-loop feedback mechanism automatically identifies which inspection items are most effective at detecting defect signs, eliminating the need for manual tracking and providing clear information about standard effectiveness through quantitative separation degree metrics.
Solution Approach 2:
The inspection standard setting system performs self-service by automatically adjusting standards based on separation degree calculations from final inspection data. The system autonomously identifies which inspection items need standard adjustments and implements them without external intervention, thereby self-tracking effectiveness through the automatic calculation and adjustment process.
Data Source
AI summary
An information processing device stores an extracted feature of each inspection item of the process inspection, and a determination result of a final inspection in a memory device, calculates a separation degree between a distribution of features of products which were determined as good products at the final inspection and a distribution of features of products which were determined as defective products at the final inspection for every inspection item or every combination of inspection items based on data of the products stored in the memory device, selects an inspection item whose inspection standard is to be reset from the inspection items or the combinations of the inspection items based on a value of the separation degree. Thus providing a method of appropriately setting an inspection standard for detecting a defect sign during process inspection. Further a process inspection device and inspection standard setting device which implements the same.