Scrap Sorting Using Vision and Analog Inductive Sensing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing scrap metal sorting methods face challenges in accurately distinguishing and separating conductive materials, particularly those with low mass, convoluted shapes, or coated wires, which often result in inefficient sorting and reduced material recovery rates.
Innovation Solution
A system utilizing a combination of vision and sensing data from an array of analog inductive proximity sensors and a control system to classify scrap materials into different categories by analyzing vision and sensing data vectors, enabling precise identification and sorting of metals and non-metals, including coated or mixed composition materials.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional sorting methods (hand sorting, air sorting, magnetic sorting) are used, then sorting can be performed on conveyor belts, but sorting accuracy is insufficient particularly for low mass, convoluted shaped, or coated wire materials
Solution Approach 1:
The patent combines multiple sensing modalities (vision system and analog proximity sensors) into a unified sorting system. The vision system captures images of scrap particles on the conveyor belt, while analog proximity sensors detect material properties through inductive coupling. These complementary data streams are integrated in a control system to achieve accurate classification of challenging materials that single-method systems cannot reliably sort.
Solution Approach 2:
The control system acts as an intermediary that processes and correlates data from both the vision system and analog proximity sensors. It reconciles the visual information about particle shape and position with the electromagnetic response data, enabling accurate identification of coated wires and convoluted shapes that would be ambiguous to either sensor type alone.
2Measurement precision
If multiple sensing systems are combined to improve sorting accuracy, then material classification improves, but device complexity increases
Solution Approach 1:
The sorting system is segmented into functionally independent modules: a vision system for capturing particle images and geometric information, analog proximity sensors for detecting electromagnetic properties, and a control system for data integration and classification. This modular segmentation allows each subsystem to be optimized independently while maintaining overall system accuracy for classifying challenging materials.
3Productivity
If high speed sorting is implemented on conveyor belts, then productivity increases, but sorting precision decreases for difficult materials
Solution Approach 1:
The vision system captures images of scrap particles before they reach the sorting decision point, allowing time for image processing and feature extraction. The analog proximity sensors continuously monitor material properties as particles pass along the conveyor. This preliminary data collection enables the control system to classify difficult materials accurately even at high conveyor speeds, maintaining both productivity and precision.
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
The system achieves improved sorting accuracy and purity, particularly for challenging materials like cast versus wrought metals, leading to enhanced material recovery and reduced contamination in sorted scrap.
Implementation Method 1
The sensing system has at least one array of analog inductive proximity sensors
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
A system and a method of sorting scrap particles is provided. A moving conveyor containing scrap particles is imaged using a vision system to create a vision image corresponding to a timed location of the conveyor, and is sensed using a sensing system to create a sensing matrix corresponding to the timed location. The sensing system has at least one array of analog proximity sensors. A control system analyzes the vision image as a vision matrix of cells, and generates a vision vector containing vision data from the vision matrix for the particle. The control system analyzes the sensing matrix, and generates a sensing data vector containing sensing data from the sensing matrix for the particle. The control system classifies the particle into one of at least two classifications of a material as a function of the vision data vector and the sensing data vector.