Event details

29. July 2026

Sensor Technology for AI-Powered Recycling Sorting Goes Into Operation

A hyperspectral camera, a metal scanner, and a color line scan camera are currently being commissioned. In the future, they will provide the data needed for robot-assisted sorting of recyclables.

At Zittau/Görlitz University of Applied Sciences, a major milestone has been reached in the RoSA-ReCycle research project: The sensor systems procured for the project have been installed at the pilot plant and are currently in the commissioning and setup phase.

The goal of the project is to automatically detect recyclable materials from heterogeneous material streams and separate them by type. To achieve this, the project team is combining several sensor principles, the data from which is then evaluated collectively. This approach makes it possible to distinguish significantly more material classes than any single method could on its own.

Three systems are being used:

  • EVK HELIOS EQ32: This hyperspectral imaging system operates in the near-infrared (NIR) spectrum and distinguishes materials based on their chemical properties—contactless and non-destructively. This allows, for example, the separation of plastic types that are visually indistinguishable.
  • EVK ABAS TL: The inductive metal scanner detects ferrous and non-ferrous metals with spatial resolution and provides the position data required for subsequent separation.
  • Teledyne DALSA Linea Color: The color line scan camera captures the shape, color, and surface characteristics of objects in the moving material stream.

The sensor package was supplied by EVK DI Kerschhaggl GmbH, a company of the Headwall Group. Julian Rössl provided on-site support for installation and training.

Close-up view of the line scan camera mounted above the material flow. It captures shape, color, and surface characteristics of the passing objects, thereby supplementing the data from the hyperspectral camera and the metal scanner.

Holger Neumann from the Institute for Process Engineering, Process Automation, and Measurement Technology (IPM) and Marc Krech from ZIRKON are responsible for the setup, integration, and parameterization. The upcoming project phases will focus on fine-tuning the sensor systems, collecting training datasets, model development, and initial sorting trials with the collaborative robot.

Co-funded by the European Union
The “RoSA-ReCycle – Robotic Sorting and Automation for Recycling” project is funded by the Just Transition Fund (JTF).
This initiative is co-financed by tax revenues based on the budget approved by the Saxon State Parliament.

Contact Person

Dipl.-Ing. (FH)
Holger Neumann
Institute of Process Engineering, Process Automation and Metrology
02763 Zittau
Theodor-Körner-Allee 8
Building Z IVc, Room C1.05
Upper floor
+49 3583 612-4447