Automated Care Area Segmentation for Semiconductor Inspection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for setting up inspection of semiconductor specimens with design and noise-based care areas are cumbersome, suboptimal, and require manual expert intervention, leading to inefficient defect detection due to limited noise data and destruction of design purity when combining care areas, resulting in suboptimal data that can bury low signal defects deep inside noise.
Innovation Solution
A system and method that utilize computer subsystems to generate, separate, and select care areas based on output attributes, allowing for automated optimization of care areas, separating instances into sub-groups with statistically different values, and adjusting inspection parameters accordingly to improve defect detection sensitivity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If manual expert intervention is used to combine care areas, then care areas can be formed for inspection, but the process is cumbersome and time-consuming (takes a week or more)
Solution Approach 1:
The system performs automated care area formation using computer subsystems that automatically generate, separate, and optimize care areas based on design and noise data without requiring manual expert intervention. The system serves itself by autonomously completing the entire care area formation process that previously required applications engineers to spend weeks manually combining care areas.
Solution Approach 2:
The manual mechanical process of expert engineers combining care areas is replaced with an automated computer-based system that uses algorithms to generate, separate, and optimize care areas. The mechanical/manual operation is substituted with computational automation that processes design and noise data to form care areas automatically.
2Ease of manufacture
If care areas are combined manually to form sensitivity regions, then inspection regions can be created, but the design purity of care area types is destroyed
Solution Approach 1:
The system separates care area instances into distinct sub-groups based on their design characteristics and noise attributes. Instead of manually combining all care areas into homogeneous sensitivity regions, the system segments them into meaningful sub-groups that preserve design purity while enabling automated inspection region creation. Each sub-group maintains its design-specific characteristics.
Solution Approach 2:
The system applies different inspection parameters and settings to different care area sub-groups based on their specific design characteristics and noise profiles. Rather than treating all care areas uniformly, each sub-group receives localized optimization tailored to its specific design purity requirements and noise characteristics, maintaining design integrity while enabling effective inspection.
3Productivity
If limited noise data is used in manual care area formation, then the process can be completed, but defect detection sensitivity is reduced and low signal defects are buried in noise
Solution Approach 1:
The system performs preliminary noise characterization and care area optimization before the actual inspection process. By pre-processing noise data and separating care areas into sub-groups based on noise attributes, the system prepares optimized inspection parameters in advance. This preliminary action enables enhanced defect detection sensitivity without requiring additional noise data collection during production inspection.
Solution Approach 2:
The system uses noise data feedback from preliminary inspections to iteratively optimize care area sub-groups and inspection parameters. By analyzing noise characteristics and adjusting care area definitions based on this feedback, the system progressively improves defect detection sensitivity even with limited initial noise data, preventing low signal defects from being buried in noise.
4Reliability
If the number of care area regions is limited to about 30, then the inspection system can operate within system constraints, but the ability to detect low signal defects is compromised
Solution Approach 1:
The system segments care areas into multiple sub-groups within each sensitivity region, allowing for more granular control and higher numbers of distinct care area definitions without exceeding system constraints. By organizing care areas hierarchically into sub-groups that can be selectively applied, the system effectively manages over 30 care area types while maintaining operational reliability and enhancing defect detection precision.
Data Source
AI summary
Methods and systems for setting up inspection of a specimen with design and noise based care areas are provided. One system includes one or more computer subsystems configured for generating a design-based care area for a specimen. The computer subsystem(s) are also configured for determining one or more output attributes for multiple instances of the care area on the specimen, and the one or more output attributes are determined from output generated by an output acquisition subsystem for the multiple instances. The computer subsystem(s) are further configured for separating the multiple instances of the care area on the specimen into different care area sub-groups such that the different care area sub-groups have statistically different values of the output attribute(s) and selecting a parameter of an inspection recipe for the specimen based on the different care area sub-groups.