Machine Learning Recipe Control for Substrate Process Drift
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Manufacturing systems face challenges in detecting and correcting substrate process drift, which leads to inconsistent substrate quality and defects due to defective components, as operators struggle to identify the cause and appropriate action to address the drift.
Innovation Solution
A system utilizing a machine learning model that processes data from sensors and metrology measurements to detect drift, predict necessary recipe modifications, and update the process recipe based on confidence criteria, identifying defective components and suggesting modifications to prevent further drift.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual monitoring and adjustment of process recipes is used, then operators can identify and address quality drift, but the complexity of identifying defective components and determining appropriate actions increases significantly
Solution Approach 1:
The system enables self-service by automatically detecting process drift, identifying defective components, and determining corrective actions without requiring operator intervention. The machine learning model autonomously monitors substrate quality parameters, compares them against target values, and generates recipe modifications when drift is detected, allowing the manufacturing system to self-correct quality deviations.
Solution Approach 2:
The patent replaces the manual mechanical system of operator monitoring and adjustment with an automated machine learning-based system. The ML model substitutes human operators in detecting drift patterns, analyzing component failures, and modifying process recipes, thereby reducing the operational complexity while maintaining or improving substrate quality consistency.
2Ease of operation
If automated machine learning-based drift detection and correction is implemented, then substrate quality consistency is maintained with reduced operational complexity, but the system complexity and computational requirements increase
Solution Approach 1:
The machine learning model serves as an intermediary between raw process data and corrective actions. It mediates by automatically translating complex metrology measurements and process parameters into meaningful drift detection and recipe modification recommendations, thereby simplifying the operator's task while managing the underlying system complexity through intelligent automation.
Solution Approach 2:
The system implements continuous feedback by monitoring substrate quality parameters in real-time, comparing them against target values, and automatically adjusting process recipes when drift is detected. This closed-loop feedback mechanism maintains substrate quality consistency while reducing operational complexity by eliminating manual monitoring and adjustment tasks.
3Manufacturing precision
If frequent process recipe modifications are made to correct drift, then substrate quality consistency is maintained, but the manufacturing productivity decreases due to increased intervention frequency
Solution Approach 1:
The system performs preliminary action by proactively detecting process drift before it significantly impacts substrate quality and automatically initiating corrective recipe modifications. This early intervention approach maintains manufacturing precision by preventing quality deviations while minimizing productivity loss by reducing the frequency and magnitude of corrections needed compared to reactive adjustment methods.
Data Source
AI summary
Methods and systems for detecting and correcting substrate process drift using machine learning are provided. Data associated with processing each of a first set of substrates at a manufacturing system according to a process recipe is provided as input to a trained machine learning model. One or more outputs are obtained from the trained machine learning model. An amount of drift of a first set of metrology measurement values for the first set of substrates from a target metrology measurement value is determined from the one or more outputs. Process recipe modification identifying one or more modifications to the process recipe is also determined. For each modification, an indication of a level of confidence that a respective modification to the process recipe satisfies a drift criterion for a second set of substrates is determined. In response to an identification of the respective modification with a level of confidence that satisfies a level of confidence criterion, the process recipe is updated based on the respective modification.