Defect Analysis Using WOE Ranking for OLED Panel Fabrication
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In the manufacturing of display panels, especially organic light emitting diode (OLED) panels, defects are difficult to trace due to the complex and integrated nature of the production process, often requiring manual data sorting based on experience, which is inefficient and unreliable.
Innovation Solution
A computer-implemented method for defect analysis that calculates weight-of-evidence (WOE) scores for device operations and contact areas to identify correlations with defects, ranking these to prioritize corrective actions, and an intelligent defect analysis system using distributed computing for predictive modeling and data visualization to quickly determine the root cause of defects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual data sorting based on experience is used for defect analysis, then flexibility in handling complex defects is maintained, but analysis efficiency and reliability deteriorate
Solution Approach 1:
The patent replaces manual mechanical data sorting with an automated computer-implemented system that uses algorithms to calculate WOE scores, rank device operations, and identify defective devices. This substitution of manual analysis with automated computational methods directly improves both efficiency and reliability of defect analysis.
Solution Approach 2:
The system enables self-service defect analysis by automatically processing fabrication data, calculating correlations between device operations and defects, and generating ranked lists of suspicious devices without requiring manual intervention. The automated system serves itself to identify root causes, improving productivity while maintaining reliability through consistent algorithmic application.
2Measurement precision
If the complete fabrication process data is analyzed to trace defects, then accuracy of root cause identification is improved, but computational complexity and data processing time worsen
Solution Approach 1:
The patent extracts only the most relevant features from the complete fabrication dataset by calculating WOE scores for each device operation and contact area. This extraction of critical information from the vast dataset improves root cause identification accuracy while reducing computational complexity by focusing only on high-correlation features rather than processing all raw data.
Solution Approach 2:
The system applies local quality analysis by evaluating each device operation and contact area individually through WOE score calculation, then ranking them by correlation strength. This localized assessment approach allows precise identification of defective devices while managing computational complexity through selective detailed analysis of only the most suspicious operations.
3Loss of information
If WOE score calculation is performed for all device operations and contact areas, then completeness of defect correlation analysis is improved, but computational time and resources worsen
Solution Approach 1:
The patent performs WOE score calculation for all device operations and contact areas (excessive action to ensure completeness), but then applies threshold-based filtering to focus only on high-correlation results. This approach maintains complete defect correlation analysis while reducing subsequent processing time by eliminating low-relevance data from further examination.
Data Source
AI summary
A computer-implemented method for defect analysis is provided. The computer-implemented method includes calculating a plurality of weight-of-evidence (WOE) scores respectively for a plurality of device operations with respect to defects occurred during a fabrication period, a higher WOE score indicating a higher correlation between a defect and a device operation; and ranking the plurality of WOE scores to obtain a list of selected device operations highly correlated with the defects occurred during the fabrication period, device operations in the list of selected device operations having a WOE score greater than a first threshold score. A respective one of the plurality of device operations is a respective device defined by a respective operation site at which the respective device perform a respective operation.