Failure Detection in CMP Substrate Processing via Recipe Comparison
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for detecting abnormalities in substrate processing apparatuses, such as CMP apparatuses, are inadequate in accuracy, particularly when changes occur in recipe data during processing, leading to incorrect determination of normal or abnormal operations.
Innovation Solution
A failure detection apparatus that collects data from sensors and compares it with recipe data stored in the system, using a determination part to identify abnormalities based on differences greater than a preset threshold, and optionally includes recipe restoration and conversion parts to adjust and verify the data for accurate detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional abnormality detection methods are used that compare sensor data with recipe data, then the detection process is simple, but the accuracy of detecting abnormality is insufficient when recipe data changes during processing
Solution Approach 1:
The system creates a backup of the original recipe data before processing begins. This preliminary action preserves the initial state, allowing later comparison with both the current recipe data and sensor readings to detect abnormalities even when recipe data changes during processing
Solution Approach 2:
The invention introduces an intermediate comparison step that contrasts sensor data against both the original recipe data and the current recipe data. This intermediary comparison mechanism enables detection of abnormalities that would be missed by conventional single-point comparison methods
2Reliability
If the system compares only current recipe data with sensor data, then the detection process is fast, but it cannot detect abnormalities caused by recipe data changes during processing
Solution Approach 1:
The system performs a preliminary backup of recipe data before processing starts. This allows the system to quickly compare sensor readings against the original recipe state without time-consuming data retrieval during detection, while still enabling reliable abnormality detection through historical reference
Solution Approach 2:
The invention creates a copy of the original recipe data that can be rapidly accessed during abnormality detection. This copying mechanism provides reliable reference data for comparison without requiring time-consuming queries to the original recipe storage system
Data Source
AI summary
The present invention improves the accuracy of detecting abnormality in a substrate processing apparatus. A failure detection apparatus 520 is provided with a data acquisition part 522 used to collect data detected by sensors 270-1 to 270-a, 370-1 to 370-b and 470-1 to470-c disposed in a CMP apparatus. In addition, the failure detection apparatus 520 is provided with a determination part 524 used to read recipe data from a recipe storage part 512, to compare the recipe data thus read with the data collected by a data acquisition part 522, thereby determining that abnormality is present in the CMP apparatus if the two data items differ.