TOMRA launches next-generation AI platform and expands GAINnext ecosystem
From reporting to interpreting: An AI agent for recycling
PolyPerception’s new AI‑agent platform marks an impressive evolution of its Waste Analyzer – an AI‑powered waste analytics solution that improves sorting performance through end‑to‑end material tracking. One of the most significant breakthroughs is the natural language interface. Operators can now 'chat' with their plant data in plain language, asking questions such as ‘How did changing the settings on the recovery line affect our purity?’. With AI as its core, the platform understands the context and provides immediate natural language answers accompanied by data breakdowns, removing the technical barrier between complex spreadsheets and operational decision-making.
While traditional AI tools in the industry are limited to 'reading' and reporting data, this platform also has 'writing' capabilities, enabling it to act like an agent within the plant. Rather than just observing material streams, it can actively create custom quality reports and set operational alerts in seconds based on its deep domain knowledge of the recycling process.
“With the introduction of our new agent-based platform, recycling plants now gain a new cognitive layer,” says Nicolas Braem, CEO and Co-Founder of PolyPerception. “Data is no longer just reported – it is interpreted, explained and transformed into relevant insights in a few seconds. Operators can interact naturally with their plant, ask questions, explore material behavior and receive clear, actionable answers in real time.”
Open data and advanced search features
This groundbreaking technology provides full transparency by allowing recyclers to integrate plant data directly into their existing management systems. This enables managers to query waste statistics or purity levels through their own dashboards without needing to log into a separate system.
The platform also introduces two powerful new search methods to help plants respond to changing material streams:
- Similarity search: Operators can right-click a problematic object, such as an electronic vape, to instantly identify every other visually similar item in the stream. This is critical for spotting fire hazards like batteries without the need to train a new AI model.
- Text and brand search: Users can search for specific brands or object types, such as 'filled refuse bags' or 'diapers', to see exactly what is passing through the facility in real time.
“AI has always been part of TOMRA’s DNA, but we are now entering an entirely new phase," says Lars Enge, EVP and Head of TOMRA Recycling. "With our acquisition of a majority stake in PolyPerception, we are moving beyond AI as a sorting tool to AI as a central intelligence for the recycling plant. By combining our advanced sorting systems and digital solutions with PolyPerception’s AI platform we are creating an end-to-end solution that doesn’t just optimize machines but fundamentally redefines how plants operate.”
Expanding the GAINnext™ ecosystem
To complement this technological progress, TOMRA is also introducing three new deep learning applications for its GAINnext™ ecosystem. This solution targets long-standing industry bottlenecks where traditional sensor-based sorting has reached its limits.
The first application addresses the rising demand for food-grade PET trays as tray material is becoming a critical new feedstock alongside bottles. By training GAINnext™ on thousands of images, the system can now distinguish between takeaway or supermarket trays and consumer or medical packaging based on shape and use. This breakthrough achieves purity levels over 95%, demonstrating that PET tray sorting is no longer a technical challenge but a viable business case.
In the metals sector, TOMRA is launching a high-precision application for 'copper meatballs', supporting a steel market that is starting its journey towards decarbonization. The new GAINnext™ automatically identifies complex copper-steel composites, such as motor armatures, even in oxidized or dirty streams, delivering outstanding selectivity and helping recyclers to upgrade rebar-grade scrap to premium furnace feedstock.
The third addition is a high-throughput solution for used beverage can (UBC) aluminum recovery from packaging streams – an application that was successfully launched in North America and has now been adapted for the European market. The GAINnext™ UBC application offers up to 33 times more throughput than manual sorting, delivering 98% purity or higher. By instantly detecting and ejecting non-UBC materials, the system provides a more efficient, automated path for aluminum can-to-can recycling.
Technology turning point
"These launches signal a true technology turning point for the industry," Enge concludes. "Deep learning is no longer just enhancing individual processes or tackling increasingly complex sorting challenges – it is linking insights directly to action across the plant. We are moving beyond high-speed detection toward a new era of intelligent, connected sorting, where complex challenges are solved and data is understood, contextualized and communicated directly to the operator. Once again, TOMRA is at the forefront of innovation, translating today’s most advanced AI into real, measurable value for customers."
