Aluminum Alloy Sorting via Machine Learning Vision
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for sorting aluminum alloys, such as x-ray transmission technology, are not cost-effective and do not accurately distinguish between cast and wrought aluminum alloys due to density variations, leading to incorrect classification and contamination in recycled molten mixtures.
Innovation Solution
A vision system with machine learning capabilities is used to identify and classify aluminum scrap pieces based on physical characteristics like color, shape, and composition, allowing for accurate sorting of cast, extruded, and wrought aluminum alloys without relying on density measurements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If x-ray transmission technology is used to sort aluminum alloys, then sorting can be performed, but the method is not cost-effective and does not accurately distinguish between cast and wrought aluminum alloys due to density variations
Solution Approach 1:
The patent replaces the mechanical/x-ray transmission-based sorting system with a vision system using machine learning algorithms. The vision system captures images of aluminum scrap pieces and uses trained neural networks to classify them as cast, extruded, or wrought based on visual features, eliminating the need for expensive and inaccurate x-ray transmission equipment while achieving superior accuracy.
Solution Approach 2:
The patent changes the measurement parameters from density-based detection (x-ray transmission) to visual feature-based detection (color, shape, texture, surface characteristics). This parameter transformation allows the machine learning model to accurately distinguish between aluminum alloy types without relying on density measurements that fail to differentiate cast from wrought alloys.
2Productivity
If density measurements are used to classify aluminum alloys, then sorting can be performed, but incorrect classification occurs due to density variations in cast and wrought alloys
Solution Approach 1:
The patent replaces the density measurement system with a vision-based machine learning system. Instead of measuring physical density properties that vary within alloy types, the system uses visual特征 (color, shape, surface texture, geometric patterns) that are consistent and distinctive for each aluminum alloy category, enabling accurate classification.
Solution Approach 2:
The patent introduces visual features as an intermediary between the aluminum scrap pieces and the classification decision. The machine learning model learns to associate specific visual patterns with each alloy type, creating a reliable intermediary pathway for classification that bypasses the unreliable density measurement approach.
3Productivity
If mixed aluminum alloys are recycled without accurate sorting, then recycling process can proceed, but contamination occurs in recycled molten mixtures reducing quality
Solution Approach 1:
The patent performs preliminary classification of aluminum scrap pieces into cast, extruded, and wrought categories before the recycling process. The vision system sorts the mixed aluminum alloys into separate streams in advance, ensuring that each category is processed separately, which prevents contamination in the recycled molten mixtures and maintains high quality of final products.
Solution Approach 2:
The patent segments the mixed aluminum alloy stream into distinct categories (cast, extruded, wrought) based on visual classification. This segmentation separates materials that would otherwise contaminate each other during recycling, allowing each segment to be processed independently and maintaining the reliability and quality of recycled aluminum products.
Data Source
AI summary
A material sorting system sorts materials utilizing a vision system that implements a machine learning system in order to identify or classify each of the materials, which are then sorted into separate groups based on such an identification or classification determining that the materials are composed of either wrought aluminum, extruded aluminum, or cast aluminum.