"The company uses an AI model that identifies different grades of aluminum based on data from lasers, X-ray fluorescence, and high-speed cameras. The system has to classify each chip - about the size of a large potato chip - in a fraction of a second. "Ten milliseconds is a long time," Siemer says. Once the vision system identifies the grade, a series of nozzles blow precise puffs of air to flip the chip off the belt and into the correct bin."
"That speed and accuracy matters because other recycling operations must melt the aluminum first before they can tell which type of alloy it is. And if alloys aren't sorted properly, the mixed heap is worth far less because customers can't be confident it will have the properties they need. "People have been wanting to go after [this unsorted aluminum], and nobody's been able to unlock it," says Siemer."
Aluminum can be recycled indefinitely but only about a third of U.S. aluminum is recycled due to difficulties in sorting mixed scrap. Sortera developed an AI-driven sorting system that identifies alloy grades using lasers, X-ray fluorescence, and high-speed cameras and classifies individual chips in milliseconds. Automated air-nozzle actuators then divert each chip into the correct bin. Accurate pre-melt sorting preserves alloy value and enables higher-priced, specification-grade outputs. Achieving accuracy above 95% unlocks previously unusable feedstock, dramatically increases margins above 90%, and enabled cash-flow-positive operations from a single plant while expansion plans proceed.
Read at TechCrunch
Unable to calculate read time
Collection
[
|
...
]