DEVELOPMENT AND FUNCTIONALITY OF THE POLYMARK INDUSTRIAL SCALE SORTING SYSTEM Hans Eder Sesotec GmbH Polymark Training & Webinar 15 March 2017 Petcore Europe Premises, Brussels
Key Topics Basics of the marker detection system Lab scale validation unit Industrial scale sorting system Performance of the sorting machine Summary
Marker Detection Basics The marker detection setup is built from two basic units A high energy UV light unit for excitation of the marker A highly sensitive camera to detect the weak fluorescence signals emitted from the marker
The UV Light Unit Equipped with high power UV LEDs at 365nm for excitation and detection of marked objects Additional visible spectra used (LEDs at Red, Green, Blue) to detect the non marked objects Optical mirrors focusing the UV energy to a line with 50mW/cm²
The Camera Unit Sensitivity > 100 times compared to standard cameras Scan rate up to 2000 scans per second Resolution of 5mm per pixel Optical long pass filter to block UV and blue fluorescence light
Lab Scale Validation Setup To validate the Polymark detection system, a small scale prototype has been developed as a first step to a sorting machine Limited in throughput (small size, low speed) Aiming to validate the UV light unit as well as the camera unit Verifying the modules fulfil all requirements for the industrial scale sorting machine Fully functional for detecting and even ejecting marked bottles
Images from Lab Scale Setup Images from trials with PET bottles passing the lab scale setup at a low speed of 10 cm/s UV light unit switched off In normal daylight there is no visible difference in marked and non marked bottles. UV light unit switched on and with dimmed ambient light. Marked bottles showing a greenish fluorescent signal. Non marked PET bottles are less bright and appear bluish.
Industrial Scale Sorting System The lab validation setup has been successfully scaled to fit on industrial scale sorting machines, with these design goals in mind: Working width of at least 1 m at a speed of 3 m/s Spatial resolution approximately 5 mm/pixel Modular platform concept, the detection system can be integrated in existing sorting machines Easily scalable to other sorting machine sizes, the working width can be extended in steps of 50 cm
Industrial Scale Sorting System Design of the industrial scale sorting machine with a platform concept for the Polymark Detection Unit.
Industrial Scale Sorting System Polymark Detection Unit mounted on the platform of the sorting machine
Industrial Scale Sorting System Inside the Polymark Detection Unit: Highly sensitive cameras High energy UV light system
Industrial Scale Sorting System Inside the Polymark Detection Unit: Every camera is covering 50cm working width UV light system can be scaled easily in steps of 50 cm
Industrial Scale Sorting System
Industrial Scale Sorting System
Industrial Scale Sorting System
Industrial Scale Sorting System
Industrial Scale Sorting System
Industrial Scale Sorting System
Industrial Scale Sorting System Industrial scale sorting machine with 1 meter working width and a speed of 3 m/s to accommodate the Polymark Detection Unit
Industrial Scale Sorting System The Polymark Detection Unit mounted on the sorting machine
Industrial Scale Sorting System View into the Polymark Detection Unit on top of the sorting machine
Performance of the Sorting Machine Setup for the Polymark sorting machine trials: Trials have been carried out on the demo recycling line of Sesotec Test setup emulated the conditions of industrial recycling lines Speed of 3 m/s on a working width of 1m at a throughput of 2 t/h
Performance of the Sorting Machine Sesotec demo recyling line with the Polymark Sorting Machine
Performance of the Sorting Machine Input material for the Polymark sorting machine trials: PET bottles provided by HERI.. fully coated with the marker partly coated with the marker label coated with the marker non coated bottles PET bottles provided by Recycling plants.. random mix of European plastic bottle collections The PET bottles from HERI have been crushed, scratched and treated with dust to emulate the condition on arrival at a recycling plant.
Performance of the Sorting Machine Examples of input material for the Polymark sorting machine trials:
Performance of the Sorting Machine General test procedure: Different mixtures of PET recycling material and coated bottles from HERI were used for the performance tests The (major) or dominating fraction in the input has always been passed through the sorting machine, the other (minor) fraction has been ejected The machine settings were optimized in order to achieve >97% purity on the major fraction with a minimal sorting loss Output purity, sorting loss, eject rate and false rejects were monitored for the different trails and are based on evaluating at least 1000 bottles
Performance of the Sorting Machine (B) Trials run with partly coated HERI Samples 80% Non Coated, 20% Partly Coated (ejected) Input Pass Eject 20 3 10 80 97 90 Eject Rate 74% False Reject 1%
Performance of the Sorting Machine (B) Trials run with partly coated HERI Samples 20% Non Coated (ejected), 80% Partly Coated Input Pass Eject 20 3 33 80 97 67 Eject Rate 89% False Reject 11%
Performance of the Sorting Machine (C) Trials run with recycling material and fully coated HERI Samples 80% Recycling Material, 20% Fully Coated (ejected) Input Eject 2 Pass 11 20 80 98 89 Eject Rate 93% False Reject 3%
Performance of the Sorting Machine (C) Trials run with recycling material and fully coated HERI Samples 20% Recycling Material (ejected), 80% Fully Coated Input 0.4 Pass Eject 20 3 80 99.6 97 Eject Rate 98% False Reject 1%
Performance of the Sorting Machine (D) Trials run with recycling material and partly coated HERI Samples 80% Recycling Material, 20% Partly Coated (ejected) Input Pass Eject 20 6 10 80 94 90 Eject Rate 77% False Reject 2%
Performance of the Sorting Machine (D) Trials run with recycling material and partly coated HERI Samples 20% Recycling Material (ejected), 80% Partly Coated Input 0.5 Pass Eject 20 21 80 99.5 79 Eject Rate 98% False Reject 7%
Summary The Polymark sorting machine is able to achieve an output purity of 98% on the major input fraction (average over all trials performed) The purity can go down as low as 94% when the bottles are only partly coated and when the input material is very diverse (as recycling material typically is). The False Rejects were at 5% on average over all trials. The Ejection Rate was at 90% on average over all trials.