Wafer Carrier Optical Detection for Tilted Placement Verification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Visual inspections for detecting tilted wafers on carriers are unreliable, leading to potential errors in placement verification during manual or robotic handling.
Innovation Solution
An article detection device comprising a lighting device, a high-definition camera, and a processor that uses a pre-trained deep neural network model to capture images and determine if wafers are correctly placed or tilted, employing a light beam to illuminate the carrier and a camera to capture images, which are then processed to detect placement situations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If visual inspection is used to detect tilted wafers, then the operation is simple and quick, but the detection reliability is poor
Solution Approach 1:
The patent replaces the manual visual inspection system with an automated optical detection system comprising a camera, lighting device, and image processing unit. This substitution transforms the mechanical/manual inspection process into an automated optical-mechanical system, achieving both improved reliability through automated image analysis and controlled complexity through standardized hardware components.
2Measurement precision
If automated image processing is implemented, then the detection precision is improved, but the device complexity increases
Solution Approach 1:
The patent replaces complex manual assessment with automated image processing algorithms that analyze captured images to determine wafer placement status. The system uses computer vision techniques to automatically identify tilted wafers, achieving high measurement precision while managing complexity through algorithmic automation rather than mechanical complexity.
Solution Approach 2:
The system creates a digital copy (image) of the wafer placement state and processes this copy to determine placement accuracy. By working with image data rather than directly manipulating physical wafers, the system achieves precise measurement without the complexity of direct physical measurement instruments.
3Measurement precision
If deep neural network model is used for detection, then the detection accuracy is enhanced, but the processing time and computational resources increase
Solution Approach 1:
The patent employs a pre-trained deep neural network model that has already learned wafer placement patterns during an offline training phase. This preliminary action of pre-training allows the model to perform rapid inference during actual detection, achieving high accuracy without the time cost of real-time training or complex processing during the detection phase itself.
Data Source
AI summary
An article detection method applied to an article detection device is provided. The method includes controlling at least one lighting device to emit a light beam towards a carrier. Once at least one camera is controlled to capture at least one image of the carrier; and a placement situation of articles in the carrier is detected based on the captured image using an article detection model.