Post by TOMRA Recycling

From optical sorting to multi-sensor intelligence. After the introduction of the first AUTOSORT™ marked the beginning of automated, sensor-based sorting in recycling, the next major steps were about flexibility, precision and control. ➡️ 2002: Machine learning unlocks system flexibility With the growth of hyperspectral imaging and more flexible software configurations, AUTOSORT™ became fully adaptable. For the first time, recyclers could adjust the system to new and varied material streams through software, without modifying the hardware. This was a major shift. AUTOSORT™ was no longer limited to one dedicated sorting task. It became an adaptable, multifunctional sorter capable of processing a broader range of materials, including plastics, paper and organic waste. ➡️ 2012: Multi-sensor intelligence & FLYING BEAM™  enable a new level of sorting precision A decade later, AUTOSORT™ took another important step forward. With the addition of an electromagnetic sensor, the platform expanded beyond NIR and VIS detection and became a true multi-sensor system. Combined with enhanced AI classification, AUTOSORT™ could interpret multiple data signals more intelligently, improving the recognition of complex material compounds and enabling more accurate separation. At the same time, TOMRA’s patented FLYING BEAM™ technology introduced homogeneous, energy-efficient illumination, strengthening NIR detection, improving sorting stability and enabling greater optical precision. Together, these milestones show how AUTOSORT™ became more flexible, more precise and more intelligent, setting the groundwork for the next stages of AI-powered sorting in recycling. Read more about the AI journey in recycling via the link in the comments ⬇️