Substrate Processing Time-Series Evaluation for Abnormality Trends
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In substrate processing apparatuses, operators face difficulty in quickly ascertaining the state of multiple processing units due to the complexity of analyzing time-series data for detecting abnormalities, as existing methods do not effectively visualize the occurrence rates and temporal changes of abnormal data.
Innovation Solution
A data processing method that calculates evaluated values by comparing time-series data with reference data, classifies these values into levels, and displays an evaluation result screen featuring graphs that visually highlight the occurrence rates and temporal changes of abnormal levels, allowing for easier identification of problematic processing units.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If all abnormalities of time-series data are displayed on the evaluation result screen, then complete information is provided, but it becomes difficult for users to easily ascertain the state of the substrate processing apparatus
Solution Approach 1:
The patent segments the evaluation result screen into multiple regions: a first graph showing occurrence rates of evaluated values across all processing units, and a second graph showing temporal changes in the number of occurrences of abnormal levels. This segmentation allows users to separately analyze distribution patterns and temporal trends without being overwhelmed by all data at once, thus maintaining information completeness while improving ease of operation.
2Measurement precision
If detailed time-series data analysis is performed for all processing units, then accurate abnormality detection is achieved, but the complexity of data analysis increases
Solution Approach 1:
The patent introduces an evaluation value calculation mechanism that compares time-series data with reference data to generate evaluated values, which are then classified into multiple levels. This intermediary evaluation system simplifies the complex time-series data into manageable categories (levels) that can be easily visualized and analyzed in the graphs, maintaining abnormality detection accuracy while reducing analysis complexity.
3Ease of operation
If occurrence rates and temporal changes are visualized using graphs, then ease of state recognition is improved, but the device complexity increases
Solution Approach 1:
The patent employs color-coded levels in the graphs where different colors represent different levels of evaluated values. This visual encoding allows users to quickly recognize the state of processing units and temporal changes without complex analysis, improving ease of operation. The color scheme is systematically designed to intuitively convey information about abnormality severity and trends.
Data Source
AI summary
A data processing method includes a step of obtaining scores of time-series data by comparing the time-series data with reference data in order to process time-series data acquired in a substrate processing apparatus having one or more processing units, a step of classifying the scores into a plurality of levels, and a step of displaying an evaluation result screen including a graph showing an occurrence rate of each level of the scores, the number of occurrences of each level, and a graph showing temporal change in the number of occurrences of a worst level of the scores when substrates have been processed through a predetermined method with respect to the processing units. Accordingly, a data processing method through which a state of the substrate processing apparatus can be easily ascertained is provided.