TOMRA Recycling, a global leader in sensor-based sorting solutions, will mark 30 years of its AUTOSORT™ platform and showcase its latest AI developments at RWM Expo 2026, taking place at NEC Birmingham on 16-17 September. At stand (ME-E274), visitors can learn about new deep learning applications for GAINnext™ and a recently launched AI platform from PolyPerception, which TOMRA is using to help UK customers improve sorting performance and support their Environment Agency compliance reporting.
With 2026 marking the 30th anniversary of AUTOSORT™, TOMRA’s multifunctional sensor-based sorting system, RWM Expo is an opportunity to reflect on three decades of continuous investment and technology enhancement in the industry-leading system. Since the first unit launched in 1996, TOMRA has developed AUTOSORT™ in line with new legislation and the changing demands of its customers. A key step came in 2019 with GAINnext™, the first deep learning technology in the industry, which opened up new applications including food-grade plastics sorting. Visitors to the stand can hear about this evolution and how AUTOSORT™ continues to help customers achieve high throughput and purity across plastics, metals and other recovered materials.
TOMRA will also introduce visitors to three new deep learning applications recently added to its GAINnext™ ecosystem. A food-grade PET tray application now distinguishes between takeaway or supermarket trays and consumer or medical packaging, achieving purity levels of over 95%. In metals, a new high-precision application for ‘copper meatballs’ identifies complex copper-steel composites such as motor armatures, even in oxidised or dirty streams, helping recyclers upgrade rebar-grade scrap to premium furnace feedstock. A third addition, for used beverage can (UBC) aluminium recovery, offers up to 33 times more throughput than manual sorting, with purity of 98% or higher.
From AI-powered sorting to AI-powered waste intelligence
Experts from PolyPerception, the AI-powered waste flow monitoring company in which TOMRA holds a 51% majority stake, will join TOMRA on the stand to showcase its new platform.
PolyPerception’s new platform transforms material data into operational intelligence, enabling facilities to understand not only what is moving through their plants, but why performance changes and where opportunities for improvement exist. It lets MRF operators ask questions about their plant data in plain English, such as how a change to recovery line settings affected purity, and get an answer straight away along with the underlying data. It can also produce custom quality reports and set operational alerts automatically. Two new search tools have been added: similarity search, which lets an operator flag a problem item such as a battery or vape and find every similar object in the stream, and text and brand search, which can track specific products or materials, such as filled refuse bags, passing through the facility in real time.
The platform also automates the sampling process and produces audit-ready evidence, giving MRF operators a practical route to meeting the Environment Agency’s updated 2024 guidance. That guidance permits AI-based visual detection for sampling, provided operators can show their methodology produces representative results. MATIQ is one UK customer deploying PolyPerception’s AI platform alongside TOMRA’s sorting technology and its digital solutions – TOMRA Insight and TOMRA Local Control – at its MRF in Sheffield to give real-time visibility of material composition, recovery rates and losses.
Terry Keyworth, Regional Director UK and Ireland at TOMRA Recycling, comments: “RWM Expo is one of the best opportunities we get each year to meet UK operators face to face, and this year we’re bringing three decades of AI-driven AUTOSORT™ innovation together with the latest capabilities from our digital solutions, GAINnext™ and PolyPerception’s new AI platform. Together, they represent TOMRA’s vision for AI-powered waste intelligence: combining sorting, operational data and AI-driven analysis to give operators a clearer understanding of what’s happening across their facilities.”
While TOMRA’s AI journey began with helping recyclers identify and recover materials more accurately, the next phase goes beyond sorting alone. By combining sensor-based sorting, deep learning and real-time data intelligence, TOMRA is helping the industry move toward AI-powered waste intelligence, enabling faster decisions, greater transparency and continuous operational improvement.
Terry adds: “This means we are ideally placed to help resolve some of the industry’s toughest current challenges, particularly around Environment Agency reporting and the Materials Facility Code of Practice. MATIQ, for example, is already using our PolyPerception technology and digital solutions at their Sheffield facility to get a clearer, audit-ready picture of what’s moving through their plant, with TOMRA sorting, digital and PolyPerception analysis working together. For operators facing more regulatory scrutiny, that combination is becoming essential.”
TOMRA Collection joins the conversation on Deposit Return Scheme implementation
TOMRA Collection, the global leader in reverse vending and a fellow division of the TOMRA Group, will also be exhibiting at RWM Expo 2026, on Stand RS-B170. Sondre Henningsgard, Vice President Business Development at TOMRA, will join a session on preparing for the Deposit Return Scheme on the second day of the show, alongside Carla Brian, Public Affairs and Partnerships Director at Biffa, and Patrick Brighty, Head of Recycling Policy at the Environmental Services Association.
