Semiconductor Process Inspection Using Simulation-Based Defect Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Advanced semiconductor processes require efficient methods to predict and interpret results while minimizing time and computing resources, with a need for improved explainability and defect detection.
Innovation Solution
A method involving obtaining measurement and simulation values, comparing them using machine learning models, and identifying defect-causing factors through similarity analysis, utilizing tools like high-resolution cameras, optical microscopes, and neural networks to inspect semiconductor processes and devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If comprehensive process analysis is performed to improve prediction accuracy, then reliability of semiconductor device characteristics is improved, but time and computing resources are excessively consumed
Solution Approach 1:
The system performs preliminary actions by capturing process data in real-time during manufacturing and pre-processing this data into structured formats. This preliminary preparation enables faster analysis when prediction is needed, reducing the time required for comprehensive process analysis while maintaining reliability.
Solution Approach 2:
The system creates simplified copies or representations of the complex semiconductor manufacturing process through structured data models and feature extractions. These copied representations can be analyzed more quickly than the full process details, enabling rapid prediction while preserving the essential information needed for reliable characteristic assessment.
2Reliability
If comprehensive process analysis is performed to improve prediction accuracy, then reliability of semiconductor device characteristics is improved, but computing resources are excessively consumed
Solution Approach 1:
The system extracts only the most relevant features and data elements from the comprehensive process information using feature extraction techniques. By taking out only the essential components needed for prediction, the system reduces computing resource consumption while maintaining the reliability of characteristic predictions.
Solution Approach 2:
The system performs preliminary data processing and feature extraction during the manufacturing process itself, preparing data in advance for prediction tasks. This preliminary action reduces the computational burden during actual prediction operations, lowering overall computing resource consumption while preserving prediction reliability.
3Measurement precision
If traditional inspection methods are used, then manufacturing process is simple, but defect detection accuracy and explainability are insufficient
Solution Approach 1:
The system introduces an intermediary layer of process data modeling and feature extraction between the raw manufacturing data and the defect detection algorithms. This intermediary structure organizes and simplifies the data relationships, enabling accurate defect detection with improved explainability while managing system complexity through structured intermediate representations.
Data Source
AI summary
A method of inspecting a semiconductor process may include obtaining a measurement value of a semiconductor device, obtaining a simulation value of a virtual semiconductor device based on input data about a manufacturing process of the semiconductor device, wherein the input data may include a suspected defect-causing factor, comparing the measurement value with the simulation value, and providing information about a predicted defect-causing factor based on a comparison result between the measurement value and the simulation value.