Sorting plastic packages with 3D relief codes

The European Circular Economy Action Plan strives to use more recycled plastics in Fast Moving Consumer Goods and hence contribute to the more sustainable use of plastics.Mechanical recycling of well-designed, food packages can yield highly pure recycled plastics that in principle could be used to produce new food packages, but unfortunately EFSA rulings demand that the recycling feedstock is composed of guaranteed high levels of food packaging only. As this guarantee cannot be given for sorted products from EPR schemes, the circular economy stifles. It is therefore crucial to develop sorting techniques that can discriminate between previous food-use and non-food-use of wasted plastic packages. A new sorting technology is proposed that is based on a 2D print technique used for anti-counterfeiting, with curved patterns of dots, not visible by the naked eye. The patterns of dots can be recognized by reflection of light with RGB cameras and specific developed software. The technique is translated to 3D with dots in relief in the packaging material. The aim is to find dots of such a size that they are hardly visible to the naked eye, but flawlessly readable for cameras. At first, patterns with dots with a diameter of 0.5 mm and a depth of 100 micron were studied. The patterns of dots are taken up in moulds of injection moulded and thermoformed packaging of in total 6 PP and 3 PET items like cups, trays, lids, for food packaging with different characteristics like transparent, coloured, and different surface qualities, produced on various machines. Additionally, a dynamic test was executed with a tailor-made sorting machine with different conveyor belt speeds and lighting strengths. The sorting yield was over 90% with no false positive showing that 3D relief dots can be used for this purpose. Thermo forming requires more control of the production process to obtain well-readable relief codes than injection moulding. A procedure is developed to take up a 3D relief code in packaging items. In next steps sorting will be tested based on realistic waste processing scenarios.