Semiconductor Equipment Anomaly Detection via Inefficiency Indexing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing semiconductor fabrication processes are inefficient due to unnecessary non-process operations, which are difficult to detect and correct, leading to decreased equipment efficiency.
Innovation Solution
A computing system that analyzes machine learning models based on equipment history data to classify operations as process or non-process, generates training data sets, and calculates inefficiency indices to identify abnormal equipment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If machine learning models are used to analyze equipment operations and detect abnormalities, then detection precision is improved, but device complexity increases
Solution Approach 1:
The system segments the complex analysis task into distinct components: operation classification (process vs. non-process), training data set generation, machine learning model development, and abnormality detection. Each component handles a specific aspect of the analysis, making the overall system more manageable and interpretable while maintaining high detection precision
Solution Approach 2:
The patent introduces intermediary elements including classification models that categorize operations, training data sets that bridge raw equipment data and analysis, and feature importance sets that mediate between complex model outputs and actionable insights. These intermediaries simplify the interpretation of machine learning results
2Productivity
If equipment operates continuously without non-process operations, then productivity is improved, but reliability deteriorates due to undetected abnormalities
Solution Approach 1:
The system implements continuous feedback by monitoring equipment operations, comparing actual performance against learned patterns from training data, and identifying deviations that indicate abnormalities. This feedback mechanism enables real-time detection of issues while maintaining continuous operation, preventing both unnecessary stoppages and undetected failures
Solution Approach 2:
The patent applies preliminary action by training machine learning models with historical equipment data before deployment. The models learn normal operation patterns in advance, enabling them to quickly identify abnormalities when they occur during continuous fabrication operations without interrupting the production flow
Data Source
AI summary
Provided is a method, performed by a computing system, of detecting an abnormality, the method including generating, a plurality of training data sets from respective history data of the plurality of pieces of equipment, generating, based on the plurality of training data sets, a plurality of machine learning models respectively corresponding to the plurality of pieces of equipment, determining, based on a plurality of feature importance sets respectively corresponding to the plurality of machine learning models, a plurality of inefficiency indices respectively corresponding to the plurality of pieces of equipment, and identifying, based on the plurality of inefficiency indices, at least one piece of abnormal equipment among the plurality of pieces of equipment.