AI brings PET trays back into the food cycle
In Duiven, the Netherlands, Cirrec, a recycling company belonging to the Faerch Group, operates a plant that recycles used PET trays from household waste for use in new food packaging. It can process up to 60,000 tonnes – or more than three billion trays – per year. AI-supported sorting technology from Steinert, a company specialising in sorting and sensor technology, plays a key role in this process.
At a glance
- Plant: Cirrec recycles PET trays in Duiven.
- Capacity: Up to 60,000 tonnes can be processed annually.
- Sorting: NIR, camera and AI achieve a purity of over 95 per cent.
- Recycled content: New packaging contains an average of around 70 per cent recycled material.
- Carbon footprint: According to Steinert, recycled PET produces 57 per cent fewer emissions than new PET.
Same plastic, different uses
PET stands for polyethylene terephthalate and is frequently used for food packaging. However, for so-called food-grade recycling – that is, processing the material so that it can be reused for food packaging – it is not enough simply to recognise the type of plastic.
The reason: a food tray and, for example, blister packaging for screws can both be made of PET. Conventional NIR sorting systems cannot distinguish between them chemically. For the food cycle, however, it is crucial to know whether a piece of packaging has actually been in contact with food previously.
For the recycling of PET trays, therefore, it is necessary to identify not only the type of plastic but also the previous use of the packaging.
AI complements sorting technology and material recognition
Cirrec uses three sorting systems from Steinert for this purpose. They combine material recognition using near-infrared (NIR) technology with image data from a colour camera and AI-based analysis. The technology helps to identify which PET packaging has actually been in contact with food and can be returned to the food cycle.
NIR technology identifies the chemical composition of the material and can thus distinguish PET from other plastics. The camera provides additional information on the appearance of the packaging. The AI analyses, amongst other things, the shape, printing and surface structure.
This enables the system, for example, to distinguish between a PET tray for a ready meal and PET packaging for a non-food product. According to Steinert, the sorting process achieves a purity of more than 95 per cent.
From used trays to new packaging
After sorting, the suitable PET is processed further. This involves, amongst other things, grinding and washing the material, which is then processed into flakes – small pieces of plastic – and pellets – small plastic granules. These serve as raw materials for new packaging produced by the Faerch Group.
On average, the new packaging contains around 70 per cent recycled material. According to Steinert, the recycled PET also generates 57 per cent fewer CO₂ emissions than newly produced PET.
In this way, used PET trays are transformed back into material for new food packaging. AI-assisted sorting is a key part of the recycling process: it helps to specifically separate suitable food packaging from other PET products, thereby closing the loop.
Key Takeaway:
For food-safe PET recycling, the sorting process must recognise more than just the plastic itself. The combination of sensors, cameras and AI makes it possible to specifically filter out suitable food packaging and reuse it to make new trays.
