Semiconductor Chamber Waveform Monitoring for Pressure Abnormality Causes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining the cause of pressure abnormalities in semiconductor manufacturing chambers are inefficient and lack precision, relying on manual pressure adjustments and waveform analysis that do not adequately account for variations in sensor data over time.
Innovation Solution
An information processing apparatus that utilizes a sensor data acquisition unit, monitoring band calculation unit, and abnormality sign detection unit to analyze sensor waveform data, calculating monitoring bands and detecting abnormalities using correlation and factor estimation rules to identify the cause of pressure fluctuations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional pressure check methods are used to determine the cause of pressure abnormalities, then the pressure abnormality cause can be identified, but real-time detection capability and accuracy in identifying time-related deviations are insufficient
Solution Approach 1:
The system performs preliminary actions by calculating monitoring bands from sensor waveform data of a predetermined number of normal runs before actual monitoring. These pre-established bands serve as reference thresholds for real-time abnormality detection, enabling the system to quickly compare current sensor data against historical norms without performing complex analysis during critical monitoring moments.
Solution Approach 2:
The system creates a copy of normal operational patterns by collecting and analyzing sensor waveform data from multiple normal runs. This copied data is then used to establish monitoring bands that represent typical behavior, allowing the system to detect deviations by comparing current readings against these replicated normal patterns rather than relying on single-point pressure checks.
2Measurement precision
If monitoring bands are calculated from multiple sensor waveform data to improve detection accuracy, then abnormality identification precision improves, but data processing complexity and computational requirements increase
Solution Approach 1:
The system segments the complex task of abnormality detection into distinct phases: data collection from multiple runs, monitoring band calculation from aggregated data, and real-time comparison against established bands. This segmentation allows each phase to be optimized independently, reducing overall computational complexity while maintaining high detection accuracy through systematic processing.
Solution Approach 2:
The system changes parameters by transforming raw sensor waveform data into derived monitoring bands that capture essential operational characteristics. By converting multiple data points into consolidated band parameters (such as mean and deviation values), the system reduces data dimensionality and complexity while preserving the information needed for accurate abnormality detection.
Data Source
AI summary
An information processing apparatus detects an abnormality sign in a semiconductor manufacturing apparatus. The apparatus includes: a sensor data collector configured to acquire sensor waveform data represented with respect to a sensor value axis and a time axis measured by a semiconductor manufacturing apparatus that is executing a process according to a same recipe; a monitoring band calculator configured to calculate each monitoring band represented with respect to the sensor value axis and the time axis used in a waveform monitoring method from a predetermined number or more of the sensor waveform data; and an abnormality sign detector configured to monitor a waveform of the sensor waveform data using each monitoring band represented with respect to the sensor value axis and the time axis and detect an abnormality sign of the semiconductor manufacturing apparatus.