Causal Wafer Polishing Control Under Environmental Variability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for determining control settings in semiconductor fabrication plants either rely on modeling-based approaches that passively observe data or active control methods like bandit experiments, which are inefficient and do not effectively account for uncontrollable environmental factors affecting wafer quality.
Innovation Solution
A control system that employs a causal model to repeatedly select and measure the impact of different control settings on semiconductor wafer quality, updating the model to accurately determine relationships between settings and environmental responses, including both controllable and uncontrollable factors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If modeling-based techniques are used to determine control settings, then the system can passively observe historical data and learn patterns, but the techniques fail to effectively account for uncontrollable environmental factors affecting wafer quality
Solution Approach 1:
The patent introduces an intermediary variable representation that separates controllable and uncontrollable factors. The causal model uses this intermediary structure to explicitly model environmental factors that cannot be directly controlled, allowing the system to account for these factors without requiring active control over them. This resolves the contradiction by enabling the system to capture environmental information while maintaining reliable control decisions.
Solution Approach 2:
The patent transforms the control approach by changing from traditional modeling parameters to causal parameters that explicitly represent environmental factors. The causal model uses counterfactual reasoning to estimate what would happen under different settings, allowing the system to account for environmental variability through parameter transformations rather than direct control.
2Measurement precision
If active control techniques like bandit experiments are used, then the system can actively control the environment for knowledge generation, but the techniques are inefficient and do not effectively determine optimal control settings
Solution Approach 1:
The patent applies preliminary action by pre-specifying the structure of the causal model before experimentation begins. The causal graph and variable representations are established in advance, which guides the experimentation process and allows the system to efficiently determine optimal control settings without requiring extensive trial-and-error experiments. This resolves the contradiction by enabling precise control settings determination through structured preliminary modeling.
3Adaptability or versatility
If traditional control methods are used, then the system can maintain stable operation, but the system cannot swiftly adapt to environmental changes and optimize wafer quality
Solution Approach 1:
The patent introduces dynamics by making the control system adaptive through causal modeling. The system continuously updates its understanding of causal relationships between control settings and environmental factors, allowing it to dynamically adjust to environmental changes while maintaining manufacturing precision. The causal model structure enables the system to learn and adapt without sacrificing control stability.
Data Source
AI summary
Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for optimizing a process of polishing semiconductor wafers. In one aspect, the method comprises repeatedly performing the following: i) selecting a configuration of input settings for polishing a semiconductor wafer, based on a causal model that measures current causal relationships between input settings and a quality of semiconductor wafers; ii) receiving a measure of the quality of the semiconductor wafer polished with the configuration of input settings; and iii) adjusting, based on the measure of the quality of the semiconductor wafer polished with the configuration of input settings, the causal model.