Spectral Product Inspection Using Posture-Aware Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing product quality determination using spectrophotometric analysis lacks accuracy due to variations in product posture during high-speed conveyance and mechanism aging.
Innovation Solution
A product inspection system that measures spectra of products in motion using a trained model generated by machine learning, considering various product postures and aging effects, to estimate the content of specific constituents accurately.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If spectrophotometric analysis is performed on products during high-speed conveyance, then productivity is improved, but measurement precision deteriorates due to posture variations
Solution Approach 1:
The system performs preliminary actions by capturing images of products at multiple different postures before the actual quality determination process. These pre-captured images in various orientations are stored and later used to create a comprehensive reference database, enabling accurate quality assessment regardless of the product's posture during high-speed conveyance inspection.
Solution Approach 2:
The system changes the parameter of product posture by intentionally capturing images in multiple different orientations and positions. By training the AI model with spectral data and images from varied postures, the system learns to recognize quality characteristics independent of specific orientation, thereby maintaining measurement precision during high-speed conveyance where posture varies.
2Device complexity
If traditional spectrophotometric analysis is used without considering posture variations, then device complexity is reduced, but measurement precision deteriorates
Solution Approach 1:
The system replaces complex mechanical adjustment mechanisms with an AI-based image recognition and spectral analysis system. Instead of using mechanical devices to physically adjust product posture or measurement angles, the system uses machine learning algorithms to analyze images and spectral data from various postures, achieving high precision measurement without additional mechanical complexity.
3Measurement precision
If multiple postures are considered in quality determination, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The system implements self-service by using AI and machine learning to automatically process and analyze spectral data and images from multiple postures. The trained model autonomously determines product quality without requiring manual intervention or complex mechanical adjustment mechanisms, achieving high measurement precision while keeping the operational system relatively simple.
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
Achieves high-accuracy product quality determination with high throughput by compensating for posture variations and mechanism aging, ensuring precise quality assessment.
Implementation Method 1
quality determination using a spectrophotometric analysis
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A product inspection system includes a spectrum measurer that measures a spectrum of a product being conveyed, and a quality determiner 44 that determines quality of the product on the basis of output obtained by inputting spectral data of the product measured by the spectrum measurer into a trained model generated by machine learning. The trained model is generated by machine learning using training data including spectral data measured in mutually different postures.