Processing Cycle Time-Series Segmentation for Semiconductor Feature Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In semiconductor manufacturing processes, such as atomic layer deposition (ALD) and atomic layer etching (ALE), the repetitive nature of processing cycles generates vast amounts of time series data, making it difficult to accurately extract feature values that reflect the processing state, especially when similar cycles are repeated hundreds of times, leading to low accuracy in feature value extraction.
Innovation Solution
An information processing method and apparatus that acquires time series data, calculates statistical values, divides the data into sections using Bayesian optimization, and generates representative values to improve feature value extraction accuracy, enabling more precise prediction and abnormality detection in repetitive processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If repetitive processing cycles are executed to manufacture semiconductor devices, then productivity is improved, but measurement precision deteriorates due to difficulty in extracting accurate feature values from vast amounts of repetitive time series data
Solution Approach 1:
The patent divides the repetitive processing cycle into multiple sections (e.g., first section, second section, third section) and extracts feature values from each section separately. This segmentation allows the system to identify characteristic patterns in different phases of the processing cycle, improving measurement precision by focusing on specific temporal regions rather than treating all repetitive data uniformly.
Solution Approach 2:
The patent extracts specific feature values from time series data at predetermined timings within each processing cycle. By selectively extracting relevant features (such as peak values, average values, or values at specific time points) rather than processing all raw data, the system maintains measurement precision while managing the complexity of repetitive processing data effectively.
2Measurement precision
If statistical processing is applied to repetitive time series data, then measurement precision is improved, but device complexity increases due to multiple calculation steps
Solution Approach 1:
The patent performs preliminary division of the processing cycle into sections and preliminary extraction of feature values at predetermined timings before final statistical analysis. This preliminary action organizes the data structure in advance, making subsequent statistical processing more efficient and reducing the overall computational complexity while maintaining measurement precision.
Solution Approach 2:
The patent applies different processing approaches to different sections of the time series data. By dividing the processing cycle into distinct sections and extracting features specific to each section, the system tailors the analysis to local characteristics of the data, improving measurement precision without requiring complex uniform processing of the entire dataset.
Data Source
AI summary
An information processing method acquires a time series data group measured during a processing cycle for a substrate. The information processing method calculates a statistical value in each cycle of the processing cycle for each of time series data included in the acquired time series data group. The information processing method generates statistical data based on the calculated statistical value. The information processing method divides the generated statistical data or time series data into predetermined sections. The information processing method calculates a representative value for each section based on the divided statistical data or time series data.