Sensor Parameter Selection for Substrate Process Result Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for analyzing large amounts of statistical data from substrate processing apparatuses are inefficient, requiring months to identify parameters significantly affecting substrate processing results, making it difficult to easily select parameters with a large effect on the outcome.
Innovation Solution
A parameter selection method involving a computer-based process that acquires sensor data, classifies parameters using specific clustering methods, selects parameters with a large effect through a tournament format, and performs correlation analysis to identify highly correlated parameters, thereby efficiently determining parameters impacting substrate processing results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If various statistical processes are performed on a plurality of sensors for each step, then comprehensive analysis of substrate processing data is achieved, but the amount of data to be handled becomes excessively large (exceeding 1,000,000 data points)
Solution Approach 1:
The patent divides the large volume of sensor data into multiple groups based on process steps. Each group contains data from sensors relevant to a specific process step, allowing the system to handle comprehensive data while managing complexity through structured segmentation. This enables efficient processing of over 1,000,000 data points by organizing them into manageable step-specific groups.
Solution Approach 2:
The patent extracts only the necessary parameters from the vast sensor data that are relevant to each process step. Instead of processing all sensor data uniformly, the system identifies and extracts key parameters specific to each step, significantly reducing the data volume that needs to be handled while maintaining comprehensive analysis capability.
2Measurement precision
If expert-based parameter selection is used, then accurate identification of significant parameters is achieved, but the analysis time becomes excessively long (several months)
Solution Approach 1:
The patent replaces the manual expert-based analysis process with an automated computer-based system. The computer automatically performs parameter selection using the grouped and extracted data, eliminating the need for expert intervention while maintaining accurate identification of significant parameters. This substitution reduces analysis time from several months to a much shorter automated process.
Solution Approach 2:
The patent performs preliminary data processing by grouping sensor data into process-step-specific groups and extracting key parameters before the actual parameter selection analysis. This preliminary organization of data enables the subsequent automated analysis to proceed efficiently without requiring expert intervention, significantly reducing the overall analysis time while maintaining accuracy.
3Adaptability or versatility
If all sensor parameters are analyzed equally, then comprehensive coverage is achieved, but the ability to identify parameters with large effects on substrate processing results is reduced
Solution Approach 1:
The patent applies different analysis approaches to different process steps by grouping sensor data according to process steps. Each group receives targeted parameter selection analysis appropriate to its specific process context, allowing the system to maintain comprehensive data coverage while identifying parameters with large effects more accurately for each specific process step.
Data Source
AI summary
A parameter selection method for causing a computer to execute processing steps including: (a) acquiring a plurality of parameters in measurement data of a plurality of sensors regarding a process in a substrate processing apparatus and result data of the process corresponding to the measurement data; (b) classifying the acquired parameters into a plurality of groups by a specific clustering method; (c) selecting parameters having a large effect on the result data based on a threshold value for each of the plurality of groups; (d) repeating the step of (c) in a tournament format between the groups for the parameters selected for each of the groups; and (e) selecting parameters highly correlated with the result data by correlation analysis between the parameters selected in the step of (d).