Image-Based Robot Alignment Using ROI Feature Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robotic systems in industries such as semiconductor manufacturing face challenges in detecting misalignment of components in a timely manner, leading to potential damage or inefficiency in processes.
Innovation Solution
A system utilizing a camera positioned in a fixed relationship to a first component, capturing images of both components during operation, and a controller that processes these images to identify visible features and determine if the second component is aligned with the first component based on predetermined positions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional laser scattering inspection tools or defect etching are used for slip detection, then detection accuracy is improved, but detection time is significantly increased causing delays in identifying misalignment
Solution Approach 1:
The patent replaces traditional mechanical/optical inspection methods (laser scattering tools, defect etching) with an image-based vision system using cameras and image processing algorithms. This substitution enables real-time detection of component misalignment by capturing and analyzing images of alignment features, eliminating the time-consuming nature of traditional inspection methods while maintaining detection accuracy.
Solution Approach 2:
The patent creates a visual copy (image) of the alignment features on components and processes this copy to detect misalignment. By capturing images of alignment features and comparing their positions across multiple images, the system can quickly identify slip conditions without physically contacting or delaying the manufacturing process, thus resolving the time-accuracy contradiction.
2Speed
If real-time image capture and processing is implemented for alignment monitoring, then detection speed is improved, but system complexity increases
Solution Approach 1:
The patent divides the monitoring task into segments by focusing on specific alignment features on components rather than analyzing entire component surfaces. The system captures images, identifies specific alignment features through image processing, and monitors their positions independently. This segmentation simplifies the overall system complexity while enabling real-time detection speed.
Solution Approach 2:
The patent implements a self-service approach where the system automatically captures images, processes them to identify alignment features, and detects misalignment without requiring external inspection equipment or manual intervention. The image processing algorithms automatically compare feature positions across images and generate detection results, reducing operational complexity while maintaining high detection speed.
Data Source
AI summary
A system for monitoring alignment of a second component relative to a first component includes a camera, and a controller including a processor and a nontransitory memory. The controller is configured to receive a first captured image from the camera when the second component is in a predetermined position relative to the first component, receive a selection of a region of interest (ROI) in the first captured image, identify a visible feature of the second component within the ROI of the first captured image, receive captured images from the camera during a subsequent operation, identify a second captured image when the second component is expected to be in the predetermined position relative to the first component, and determine if the second component is in the predetermined position relative to the first component based on the second captured image and the identified visible feature of the first captured image.