Tomra’s artificial intelligence (AI)-based plastics sorting has achieved a new milestone. GAINnext, the company’s deep learning sorting solution which works as an add-on for Tomra’s Autosort machines, can now quickly and efficiently separate food-grade from non-food-grade plastics for PET, PP and HDPE on a large scale.
The Norway-based sorting machine manufacturer has described the development as a ‘worldwide industry first’.
“We have used AI technology to improve sorting performance for decades, but this latest groundbreaking application marks another industry first for us,” said Dr. Volker Rehrmann, Head of Tomra Recycling. “AI has the power to transform resource recovery as we know it, and our latest sophisticated applications of deep learning and AI reinforce our position as a pioneer in this field. With its sophisticated use of deep learning, GAINnext enables food-grade sorting and bottle-to-bottle quality, tasks that have posed significant challenges for our industry for many years. The use of AI is driving material circularity at a time when it is needed most, with tightening regulations and increasing customer demand for technologically advanced solution,” he added.
Food-grade sorting is challenging because food and non-food packaging are often made of the same material and visually very similar. Hygiene concerns and increasingly stringent industry regulations add a further layer of complexity to handling food waste in recycling.
Tomra’s deep learning system is used categorise objects based on an image. Its engineers trained neural networks with thousands of images that hold a pool of object information. Based on this information, the network recognises patterns and properties and connects the information pool to the sorting task, classifying the objects transported on the conveyor belt and scanned by near-infrared, visual spectrometry, and other sensors.
The AI-based sorting solution is achieving purity levels over 95% for the packaging applications in plants across UK and Europe, Tomra said in a statement.
Tomra first introduced its deep learning technology in 2019 with an application to identify and remove PE-silicon cartridges from polyethylene streams.
To date, more than 100 Autosort units with GAINnext are installed at material recovery facilities across the globe. Users include Berry Circular Polymers’ facility in Leamington Spa, Viridor Avonmouth in Bristol, and the French Nord Pal Plast plant, owned by the European Dentis Group.