Physics based AI unlocks first global predictions of carbon cycling in ocean sediments Researchers at The University of Manchester have developed a new physics‑based artificial intelligence approach that, for the first time, enables accurate global‑scale predictions of how dissolved organic carbon moves between seawater and marine sediments, a crucial but previously unquantifiable component of the
AI Sorting Solutions at IFAT 2026 – RECYCLING magazine
Tomra Recycling will present new AI sorting solutions at the IFAT 2026 in Munich from 4 to 8 May. At stand 339 in Hall B6, the company will mark 30 years of its AUTOSORT™ system and outline developments in sensor-based and AI-supported sorting technologies.
Development of sensor-based sorting
Since the introduction of Autosort in 1996, sorting technology has evolved from rule-based optical systems to integrated and connected solutions. Early systems combined automated sorting with optical sensors and initial AI approaches. Subsequent developments included hyperspectral imaging, improvements in illumination and detection of black materials, as well as increased flexibility through software-based configurations.
In 2019, the integration of the deep learning module GAINnext extended sorting capabilities, particularly in complex applications such as food-grade plastics. Current systems combine multi-sensor technology, image processing and AI-based material recognition with digital tools for monitoring and process optimisation.
Software integration and plant control
A new software platform will also be introduced at IFAT 2026. Tomra Local Control enables centralised operation of sorting systems via a single interface within the plant network. The system provides continuous performance data and allows remote adjustment of sorting parameters and configurations. This supports data-driven plant management and operational efficiency.
Data-driven waste analysis
At the exhibition, Tomra will also present technologies for waste analysis developed in collaboration with PolyPerception. These systems use deep learning to analyse material streams in real time, providing information on composition, recovery rates and material purity. The resulting data can support process optimisation and quality control in sorting facilities.
Expansion of AI-based applications
In addition to existing systems, Tomra will introduce further AI sorting solutions based on deep learning. These developments are intended to expand application areas and improve sorting performance across different material streams. Company representatives will provide technical information on the new systems and their integration into existing plant infrastructure.
Tomra at IFAT
Hall B6, Booth 339
Related Posts
