Sorting Copper-Containing Parts from Ferrous Scrap
Industry - Systems integration - Machine control
Automatically sorting copper-containing objects from a ferrous stream using Vision-AI and robotics.
Background
In metal recycling, material streams are mechanically separated to recover valuable raw materials. After the shredding process, ferrous parts are separated from the other materials using a magnetic roller.
This ferrous waste stream often contains objects that consist of copper as well as iron. Examples include electric motor armature cores, tangled cables, wires, and other copper-containing metal objects. Because of their ferrous properties, these objects remain in the ferrous waste stream, even though they have a valuable composition.
Riwald Recycling recycles hundreds of thousands of metric tons of waste materials each year, ranging from demolition waste and high-value industrial waste streams to train cars and airplanes. These materials are then recycled into more than 150 types of sustainable raw materials.
Riwald Recycling is looking for a solution to recover these materials automatically and reduce its reliance on manual sorting.
Challenge
Currently, manual sorters visually identify copper-containing objects and remove them manually from the material stream.
This application is characterized by a highly unstructured material flow. Objects vary continuously in shape, size, color, and position on the conveyor belt. In addition, many objects contain both ferrous and non-ferrous components. As a result, traditional detection techniques are unable to distinguish between them adequately.
The challenge is to develop a reliable sorting system that can independently identify, classify, and locate copper-containing objects within a continuous process.
Solution
RIWO is developing a smart sorting system based on Vision-AI, RGB-D/3D cameras, and AI classification. Using deep learning and machine learning, the system is trained to recognize a wide variety of copper-containing objects. The Vision-AI software continuously analyzes the material flow and determines which objects should be sorted out.
The combination of color information, depth data, and AI models makes it possible to accurately recognize, classify, and position objects. Based on this information, a FANUC robot is controlled to automatically remove the selected objects from the ferrous stream.
The result is an automated process for recovering copper-containing materials in a challenging recycling environment. The system contributes to higher material purity, improved recovery of valuable raw materials, and reduced reliance on manual labor. It also creates a safer work environment and represents a future-proof step forward in recycling automation.
Be convinced
by the power
from RIWO
Contact us for more information.