Processing Chamber Calibration via Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for calibrating substrate processing equipment are inefficient, relying on manual adjustments that are time-consuming and prone to errors, leading to non-ideal manufacturing parameters, faulty substrates, and reduced yield due to unmeasured parameters and chamber drift over time.
Innovation Solution
A method involving machine learning models that receive sensor data from multiple sensors to tune calibration parameters of physics-based models, allowing for real-time calibration and accounting for chamber-specific variations, thereby improving accuracy and consistency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If manual calibration methods are used for substrate processing equipment, then operators can adjust parameters directly, but the process becomes time-consuming and error-prone leading to reduced productivity and manufacturing precision
Solution Approach 1:
The system performs self-calibration by automatically comparing sensor readings with digital twin predictions and adjusting parameters without human intervention. The machine learning model autonomously identifies calibration drift and corrects it, eliminating the need for manual operator involvement while maintaining accuracy.
Solution Approach 2:
The patent replaces manual mechanical calibration operations with an automated digital system. Instead of operators physically adjusting equipment parameters, a machine learning-based digital twin system automatically calculates and applies calibration corrections based on sensor data and virtual model comparisons.
2Ease of operation
If manual calibration adjustments are performed, then parameter changes can be made directly, but accuracy is compromised due to unmeasured parameters and chamber drift over time
Solution Approach 1:
The system continuously monitors sensor data and compares actual readings with digital twin predictions to detect calibration drift. This feedback loop enables the system to identify when parameters have deviated from optimal values and automatically correct them, maintaining high precision over time without manual intervention.
Solution Approach 2:
The digital twin model predicts optimal calibration parameters in advance by simulating chamber behavior and comparing virtual sensor readings with actual measurements. This preliminary calculation allows the system to proactively adjust parameters before significant drift occurs, maintaining manufacturing precision.
3Measurement precision
If comprehensive sensor data collection is implemented to account for all parameters, then calibration accuracy improves, but system complexity and data processing requirements increase
Solution Approach 1:
The digital twin acts as an intermediary between multiple sensors and the calibration system. Instead of directly processing complex data from numerous sensors, the system compares sensor readings with predictions from the virtual model, simplifying the analysis while maintaining comprehensive monitoring of all relevant parameters.
Solution Approach 2:
The machine learning model serves multiple functions simultaneously: it processes data from various sensor types, detects calibration drift, predicts optimal parameters, and guides corrections. This multi-functional approach consolidates what would otherwise require separate systems for each task, reducing overall complexity.
Data Source
AI summary
A method includes receiving, from sensors, sensor data associated with processing a substrate via a processing chamber of substrate processing equipment. The sensor data includes a first subset received from one or more first sensors and a second subset received from one or more second sensors, the first subset being mapped to the second subset. The method further includes identifying model input data and model output data. The model output data is output from a physics-based model based on model input data. The method further includes training a machine learning model with data input including the first subset and the model input data, and target output data including the second subset and the model output data to tune calibration parameters of the machine learning model. The calibration parameters are to be used by the physics-based model to perform corrective actions associated with the processing chamber.