High accuracy sorting of multiple material streams with deep learning technology.

GAINnext™ maximizes the recovery and purity levels of valuable materials for recycling. Its AI-based visual classification identifies thousands of objects in milliseconds, offering high-throughput processing with up to 2000 ejections per minute, depending on the application.

Thanks to its integrated deep learning software, GAINnext™ identifies hard-to-classify objects, reducing the need for manual sorting. The system automates complex tasks that cannot be achieved with traditional optical waste sorting. The intelligent system also classifies overlapping objects and calculates trajectory positioning for high-precision sorting.

GAINnext™ makes it possible to process multiple material streams on a single sorting line. The future-proof, adaptive technology can be easily optimized with network connectivity to continually improve performance and reduce downtime. Cloud-based monitoring and digital application services are optionally available with TOMRA Insight.

What can GAINnext™ sort?

Food-grade PET bottles

Separates food-contact PET bottles from non-food containers such as detergent and spray bottles in mixed streams, enabling high-purity rPET for bottle-to-bottle recycling without manual sorting. 

Food-grade PET trays

Recovers high-value transparent, monolayer PET trays from mixed streams by separating food-contact trays from non-food applications, enabling high-purity rPET and supporting tray-to-tray recycling. 

Food-grade PP packaging

Identifies and separates food-contact PP packaging such as yogurt cups and tubs from non-food applications like cosmetics and cleaning products, enabling high-purity rPP for consistent, high-quality recycling.

Food-grade HDPE packaging

Separates food-contact HDPE containers such as milk and water bottles from non-food packaging like detergent and shampoo containers in mixed streams, enabling high-purity rHDPE for food-contact compliant recycling.

PET cleaner

Removes hard-to-classify contaminants such as opaque PET, textiles, and multilayer foils from PET streams, improving the purity of transparent and light-blue PET fractions for high-quality recycling.

PE silicone cartridges

Uses AI-based visual classification to detect and remove silicone cartridges from PE streams, producing high-purity polyethylene fractions and improving downstream recycling performance. 

Aluminum used beverage cans (UBC)

Identifies UBCs based on visual characteristics such as shape and size, separating them from contaminants like aluminum bottles, food cans, and trays to produce high-purity aluminum fractions.

Wrought aluminum upgrade

Upgrades wrought aluminum scrap by detecting and removing low-alloy cast from mixed metal streams, enabling high-purity wrought fractions with low silicon content for premium recycling applications. 

Copper meatballs

Removes copper–steel compounds such as copper meatballs and motor armatures from ferrous scrap, reducing residual copper levels and enabling higher-quality steel recycling feedstock.

Deinking / Paper cleaning

Improves deinking performance by removing hard-to-classify materials such as colored paper, envelopes, receipts, and brown board, enabling high-purity paper fractions for recycling.

Wood (A vs B, C&D)

Enables high-purity wood recycling by classifying waste wood and identifying materials such as untreated wood, engineered wood, and MDF through automated sorting and impurity removal.

One AI system. Multiple configurations.

GAINnext™ adds a new layer of visual intelligence to TOMRA’s sorting systems – enhancing material classification and enabling object recognition where it matters most. It can be integrated into existing setups or deployed as a standalone solution, with our experts defining the right configuration for your specific sorting requirements.