Spectral Product Inspection Using Posture-Aware Machine Learning

Resolve Bottlenecks,
Find Innovative Solutions
Generate 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

VSEngineering 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

Engineering Contradiction:
ImprovethroughputVSAvoidquality determination accuracy
Core Design Contradiction:
ProductivityVSMeasurement precision

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.

Inventive Principle:
Principle #10Preliminary action

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.

Inventive Principle:
Principle #35Parameter changes

2Device complexity

If traditional spectrophotometric analysis is used without considering posture variations, then device complexity is reduced, but measurement precision deteriorates

Engineering Contradiction:
Improvesystem simplicityVSAvoidquality determination accuracy
Core Design Contradiction:
Device complexityVSMeasurement precision

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.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

3Measurement precision

If multiple postures are considered in quality determination, then measurement precision is improved, but device complexity increases

Engineering Contradiction:
Improvequality determination accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

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.

Inventive Principle:
Principle #25Self-service

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

Methodology Applied
Scientific EffectSpectrophotometric analysis: Absorption Spectroscopy

Data Source

PatentEP4641177A1Product inspection system and product inspection method
Publication Date: 2025.10.29 USHIO INC
  • EP4641177A1 patent drawingFigure 1
  • EP4641177A1 patent drawingFigure 2
  • EP4641177A1 patent drawingFigure 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.