AI Vision Feed Bowl for Oriented Parts and Defect Rejection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing high-throughput feed systems for oriented objects, such as vibrating or centrifugal bowls, face challenges in rapid object changing and fail to detect defective objects, leading to inefficiencies and waste due to the need for dedicated bowls for each object geometry and potential machine stoppages.
Innovation Solution
A feed bowl system integrated with a vision system and artificial intelligence algorithms that uses a learning phase to define acceptable object orientations and qualities, enabling rapid adjustment and real-time detection of defects through image compression-decompression models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If dedicated bowls are used for each object geometry to ensure precise orientation, then manufacturing precision is improved, but device complexity and loss of time increase due to needing multiple bowls and changeover time
Solution Approach 1:
The patent implements a single multi-functional bowl that can handle multiple object geometries through software-based object recognition and adaptive orientation control. The system uses a camera to capture images of objects in the bowl, processes these images to identify object geometry and orientation, and dynamically adjusts vibration parameters to orient different object types correctly, eliminating the need for dedicated bowls for each geometry.
Solution Approach 2:
The patent replaces the mechanical approach of using physically different bowls for different objects with an electronic/software-based system. A camera captures images, a processor analyzes object geometry and orientation, and control algorithms adjust vibration parameters accordingly. This substitution of mechanical diversity with electronic adaptability resolves the contradiction between precision and complexity.
2Productivity
If adjustment time is reduced for rapid object changing, then productivity is improved, but manufacturing precision deteriorates due to insufficient time for proper bowl adjustment
Solution Approach 1:
The system performs preliminary actions by pre-programming orientation algorithms and object recognition patterns for multiple geometries. When an object change is detected, the system already has the necessary orientation parameters and vibration profiles ready, allowing immediate adaptation without manual adjustment. The camera and processor are continuously monitoring and ready to process new object types instantly.
Solution Approach 2:
The system dynamically adjusts vibration frequency, amplitude, and phase in real-time based on detected object geometry and orientation. Rather than requiring static pre-adjustment for each object type, the control system continuously adapts parameters during operation, enabling rapid changeover while maintaining precision through real-time feedback from the camera and processor.
3Manufacturing precision
If visual inspection with AI algorithms is implemented to detect defects, then manufacturing precision is improved, but device complexity and use of energy increase
Solution Approach 1:
The patent introduces a camera as an intermediary sensing device that captures images of objects in the bowl. This optical intermediary provides rich visual information about object geometry, orientation, and potential defects without requiring complex mechanical measurement systems. The camera images are then processed by AI algorithms to extract quality metrics, enabling high-precision inspection with relatively simple hardware.
Solution Approach 2:
The system creates digital copies (images) of the physical objects using a camera. These image copies are then analyzed by AI algorithms to detect defects, measure geometry, and determine orientation. This copying approach allows complex analysis to be performed on digital representations rather than requiring complex physical measurement apparatus, thereby improving precision while limiting complexity growth.
4Device complexity
If a single bowl handles multiple object geometries instead of dedicated bowls, then device complexity is reduced, but manufacturing precision deteriorates due to generic orientation settings
Solution Approach 1:
The system implements continuous feedback through a camera that monitors object position and orientation in the bowl. The processor analyzes these images to detect actual object geometry and current orientation, then feeds this information back to the vibration control system. This closed-loop feedback enables the single bowl to achieve precise orientation for different geometries by dynamically adjusting vibration parameters based on real-time object state.
Solution Approach 2:
The system achieves precision with a single bowl by dynamically changing vibration parameters (frequency, amplitude, phase) based on detected object geometry. Rather than relying on fixed mechanical settings specific to each object type, the control algorithm adjusts vibration parameters in real-time to match the physical characteristics of the current object, enabling one bowl to perform the work of multiple dedicated bowls with equal precision.
Data Source
AI summary
The method of feeding objects such as tube tops or caps comprises at least one orientation and quality inspection step integrated into the feeding method effected continuously during production, the orientation and quality inspection comprising a learning phase and a production phase.