Automatic Metrology Site Detection Using Object Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Manual definition of metrology sites for product inspection is prone to uncertainties and becomes challenging with complex measurement techniques, especially when scaling from sample analysis to large-volume production.
Innovation Solution
A method using a computing device for detecting and aligning objects in measurement data, followed by analysis to determine metrology sites, which includes object detection through template matching or machine learning, alignment into a common reference system, and visualization of variances to define measurement locations automatically or with human input.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual inspection by experienced human operators is used to define metrology sites, then expertise-based quality judgment can be applied, but scaling to huge volume production introduces uncertainties and reduces efficiency
Solution Approach 1:
The system enables automatic self-determination of metrology sites through machine learning models that analyze product images and autonomously identify measurement locations, eliminating the need for continuous human operator intervention while maintaining high accuracy across volume production
Solution Approach 2:
The patent replaces the mechanical human inspection process with an automated computer vision system using machine learning algorithms to detect and define metrology sites, substituting human expertise with an scalable automated system that processes images consistently across large production volumes
2Ease of operation
If two-dimensional images are used for manual analysis, then inspection is feasible for experienced operators, but the process becomes much more difficult when three-dimensional imaging, spectroscopy or combinations of measurements are used
Solution Approach 1:
The machine learning system is designed to handle multiple measurement modalities (2D imaging, 3D imaging, spectroscopy) through a unified framework, allowing the same automated system to process diverse measurement types without requiring separate manual analysis procedures for each modality
Solution Approach 2:
The patent introduces machine learning algorithms as an intermediary that processes complex multi-modal measurement data and translates it into standardized metrology site definitions, making the system accessible and manageable regardless of the underlying measurement technique complexity
3Productivity
If manual analysis of sample products is scaled to huge volume production, then production capacity increases, but uncertainties in metrology site definition increase
Solution Approach 1:
The system incorporates feedback mechanisms where the machine learning model continuously learns from processed images and measurement outcomes, refining its metrology site definition accuracy with each processing cycle and maintaining consistent reliability across increasing production volumes
Solution Approach 2:
The patent utilizes parameter changes in the machine learning model during training and operation to adapt to different product variations and measurement conditions, maintaining reliable and consistent metrology site definitions across diverse production scenarios and volume scales
Data Source
AI summary
Methods for determining metrology sites for products includes detecting corresponding objects in measurement data of one or more product samples, and aligning the detected objects are aligned. The methods also include analyzing the aligned objects, and determining metrology sites based on the analysis. Devices use such methods to determine metrology sites for products.