Wafer Mapping Vision Array for Fast Load Port Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current substrate mapping techniques in semiconductor fabrication equipment, such as vacuum-suction and break-beam methods, are inefficient due to complexity, high costs, and inaccuracies in detecting substrates with varied thicknesses and orientations, leading to reduced throughput and increased manufacturing costs.
Innovation Solution
A machine vision system with cameras and illuminators is integrated into the semiconductor fabrication equipment to capture and process images of substrates, determining their presence, orientation, and edge profiles, allowing for precise mapping and adjustment of end effector tines to handle warped or bowed substrates without moving parts, thus enhancing substrate handling efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If vacuum-suction technique is used to detect substrate presence, then substrate state can be determined, but the technique is slow (hundreds of milliseconds) and may give false readings when sealing is not tight
Solution Approach 1:
The patent replaces the mechanical vacuum-suction detection system with an optical imaging system using cameras and illuminators. The machine vision system captures images of substrates at holding locations and uses image processing to determine substrate presence, orientation, and state, eliminating the slow vacuum valve operation and sealing dependency while maintaining detection accuracy.
2Measurement precision
If break-beam technique is used to map substrate stack, then substrate presence can be detected, but the technique requires precision alignment and involves complex moving parts that slow down the process
Solution Approach 1:
The patent replaces the mechanical break-beam sensor system with a stationary machine vision system. Multiple cameras and illuminators are positioned to capture substrate images without requiring moving parts or precision mechanical alignment. The optical imaging system maintains detection precision while eliminating complex mechanical components and extending/retracting mechanisms.
Solution Approach 2:
The patent creates optical copies (images) of substrates using cameras and illuminators. These image copies are then processed to extract substrate position, presence, and state information, replacing the need for physical beam interruption detection and mechanical scanning of break-beam sensors.
3Loss of information
If conventional imaging systems are used to map substrates through load port opening, then substrate images can be captured, but images of top or bottom substrates may be distorted or blocked by other substrates
Solution Approach 1:
The patent segments the imaging task by using multiple cameras positioned at different locations (e.g., top, bottom, side views) to capture images of different portions of the substrate stack. Each camera captures a specific region or angle, ensuring that substrates blocked from one view are visible from another, thereby eliminating distortion and blocking issues.
Solution Approach 2:
The patent adds spatial dimensions to the imaging system by positioning cameras at multiple locations and angles around the substrate stack. Instead of a single viewpoint that suffers from occlusion, the system uses multi-dimensional positioning to ensure all substrates are visible from at least one camera angle, eliminating image blocking and distortion.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The machine vision system improves substrate mapping accuracy and throughput by eliminating the need for complex mechanical systems, accurately detecting substrate states and orientations, and enabling efficient handling of substrates with varied thicknesses and warping, reducing manufacturing costs and increasing processing efficiency.
Implementation Method 1
A machine vision system with cameras and illuminators is integrated into the semiconductor fabrication equipment to capture and process images of substrates
Data Source
AI summary
A semiconductor wafer mapping apparatus comprising a frame forming a wafer load opening communicating with a load station for a substrate carrier disposed to hold more than one wafers vertically distributed in the substrate carrier for loading through the wafer load opening, a movable arm movably mounted to the frame so as to move relative to the wafer load opening and having at least one end effector movably mounted to the movable arm to load wafers from the substrate carrier through the wafer load opening, an image acquisition system including an array of cameras arranged on a common support and each camera fixed with respect to the common support that is static with respect to each camera of the array of cameras, wherein each respective camera is positioned with a field of view disposed to view through the wafer load opening with the common support positioned by the movable arm.