IFAT 2026 : TOMRA unveils new deep learning solutions at IFAT 2026

TOMRA has led the industry transition from pioneering automated sorting to fully integrated, connected systems since the introduction of the first AUTOSORT™ 30 years ago. Visitors to the stand can learn about the evolution of the AUTOSORT™ system and the wider history of sensor- and AI-based sorting by exploring the different versions launched since its debut.

This journey began with the 1996 edition, which introduced high-capacity automated sorting supported by smart optics and the earliest principles of AI. Over the years, hyperspectral imaging and software-based flexibility have enabled adaptation to new material streams. Optical upgrades have enhanced illumination and made black material detection possible, while advanced AI-driven material intelligence has expanded accuracy and material range. A critical milestone was reached in 2019 with the integration of GAINnext™, the first deep learning technology on the market. As an add-on to the AUTOSORT™ system, GAINnext™ unlocked entirely new sorting applications and solved long-standing challenges such as sorting food-grade plastics.

Today, AUTOSORT™ is recognised as the most powerful sorting system on the market. It combines AI, multi-sensor technology and advanced image processing, and is enhanced by tools such as GAINnext™ for waste analysis, as well as other digital tools for real-time monitoring, data-driven optimisation and seamless plant integration. This makes it the connected core of high-quality recycling operations worldwide.

Lars Enge, EVP and Head of TOMRA Recycling, comments: "For three decades, the AUTOSORT™ brand has set the benchmark for performance in the recycling industry. IFAT 2026 offers us the ideal stage not only to reflect on our history, but also to share a clear vision of what comes next. We will introduce several new, impressive deep learning and AI solutions that help bridge the gap to true circularity. These innovations enable sorting plants to move beyond high-speed detection toward truly intelligent, interconnected and user-friendly ecosystems.” 

Alongside the award-winning AUTOSORT™ and GAINnext™ technologies, TOMRA’s focus for IFAT 2026 is its brand-new, intuitive software platform. TOMRA Local Control enables operators to oversee all sorting machines via a unified interface within the plant network. It provides full visibility of every machine on the line and delivers updates on machine status and performance trends every second. Operators can adjust configurations, activate sorting plans and create custom programmes remotely from the office, transforming plant control into a faster, more efficient, data-driven experience.

Visitors to the TOMRA booth will also have the opportunity to meet the PolyPerception team, whose AI-powered waste analysis technology brings a new layer of intelligence to modern sorting plants. Its deep learning waste analyser delivers full transparency and traceability across the entire line, providing real-time insights into material composition, recovery and purity. By converting continuous stream data into reliable, actionable information, PolyPerception empowers operators to enhance efficiency, optimise performance, and make informed, data-driven decisions, thereby reinforcing TOMRA's vision of a smarter, more interconnected and forward-looking recycling ecosystem.

TOMRA’s plans for the show also include the launch of various new deep learning and AI solutions that will be presented to the public for the first time. Experts will be available to discuss these latest developments in deep learning, along with the company's broader portfolio of sensor-based sorting solutions. 

“We welcome customers, partners and industry companions to our TOMRA booth in Hall 6 to discover what’s next in AI sorting,” concludes Enge.