Upgrading Zorba with Eddy Current and AI Sorting
Zorba is not a single metal. It is a shredded mixed non-ferrous scrap stream dominated by aluminum but potentially containing copper, brass, zinc, stainless steel and other metals.
That mixture creates an important distinction for recycling plants: recovering non-ferrous metal from waste and upgrading a mixed non-ferrous fraction are two different jobs.
Recent industry developments are increasingly focused on the second job. In September 2026, industry reporting highlighted advanced sorting of mixed Zorba into more differentiated secondary metal streams, reflecting growing interest in extracting additional value after primary metal recovery.
Eddy Current Separation Does the Bulk Recovery
An eddy current separator is well suited to the primary recovery of conductive non-ferrous metals from non-conductive material.
Its rapidly changing magnetic field induces electrical currents in conductive particles. The resulting force changes their discharge trajectory, allowing aluminum and other conductive metals to separate from plastics, rubber, glass and similar non-metallic material.
CurrenTek lists aluminum, copper and brass among the materials targeted by its eddy current separators. In a Zorba-related process, this technology performs the bulk recovery work rather than attempting to identify every individual alloy.
Feed preparation still matters. Particle size, moisture, material layer thickness and stable feeding can all influence separation behavior, so an eddy current separator should be configured around representative material rather than a generic scrap description.
Why Zorba May Need Further Upgrading
After primary separation, the recovered metal fraction can still contain different non-ferrous metals and difficult pieces.
This is where the recycling objective changes. The plant is no longer simply asking, “Is this material conductive?” It may instead need to identify selected pieces within an already metal-rich stream.
Recent developments in sensor-based metal sorting illustrate this trend. At ALUMINIUM 2026, for example, visible-light AI systems were being presented for recognizing non-ferrous scrap from surface characteristics. Other technologies such as XRT and LIBS are also being developed for more specialized metal and alloy classification.
These technologies should not be treated as interchangeable: they measure different material characteristics and solve different sorting problems.
Where CurrenTek AI Sorting Fits
CurrenTek describes its AI Separator as an upgrading stage after primary separation.
The system uses image recognition and intelligent classification to analyze visible characteristics of individual pieces. Published applications include selected metal fractions, wire and cable materials, stainless steel and aluminum-rich scrap, depending on whether the target material can be reliably distinguished from the feed.
That limitation is important. CurrenTek does not describe this machine as XRT, LIBS or chemical-composition analysis, so it should not be presented as an alloy analyzer.
Instead, its role is to provide an additional recognition step where target pieces have usable visual differences.
Design the Process Around the Product
For a Zorba processor, the correct question is not whether AI should replace eddy current separation. The two stages address different tasks.
A practical process may use screening for consistent sizing, magnetic separation for residual iron, eddy current separation for bulk conductive non-ferrous recovery, and AI sorting only where an identifiable target fraction justifies further upgrading. CurrenTek lists all four equipment categories within its Zorba recycling solutions.
The final configuration should depend on feed composition, particle size, contamination and the product the recycler intends to sell.
The commercial objective is straightforward: perform inexpensive bulk separation first, then apply more selective sorting only where it can create a useful upgraded fraction.