Etching Quality Prediction From Transfer-Path Substrate Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current etching processing systems face challenges in controlling the feedback of etching quality inspection results to subsequent substrate processing due to the time required for external inspection devices to obtain etching quality results.
Innovation Solution
An etching processing system that includes a memory and a processor configured to predict etching quality by inputting image data of substrates into a trained model, associating image data with etching quality information, allowing for real-time prediction of etching quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If external inspection devices are used to inspect etching quality, then measurement precision is improved, but loss of time increases
Solution Approach 1:
The imaging device captures images of the substrate during the transfer path before external inspection, and the trained model predicts etching quality in advance. This preliminary action allows the system to obtain etching quality results before the substrate completes the full processing cycle, reducing the time loss while maintaining measurement precision through the trained prediction model.
Solution Approach 2:
A trained prediction model serves as an intermediary between the imaging device and external inspection devices. The model processes images captured during transfer and provides predicted etching quality results, acting as a mediator that delivers accurate measurements without requiring the time-consuming external inspection process for every substrate.
2Manufacturing precision
If external inspection devices are used to obtain etching quality results, then manufacturing precision is improved, but productivity decreases
Solution Approach 1:
The system performs preliminary etching quality prediction using the trained model during the substrate transfer path, before external inspection. This allows manufacturing precision to be maintained through accurate prediction while increasing productivity by eliminating the need to wait for external inspection results before proceeding with subsequent processing of the next substrate.
Solution Approach 2:
The trained model provides real-time etching quality prediction feedback during the transfer path, enabling immediate adjustment of processing parameters for subsequent substrates. This feedback mechanism maintains manufacturing precision while improving productivity by allowing continuous processing without waiting for external inspection results.
Data Source
AI summary
An etching processing system includes a memory, and a processor coupled to the memory and configured to predict etching quality of a substrate by inputting image data of the substrate into a trained model trained by using training data in which image data of substrates captured by an imaging device arranged on a transfer path of the substrates and information for predicting etching quality of the substrates are associated with each other.