AI BriefingHow AI Helps Recyclers Turn Mixed Waste Into Reusable Material
·2026.11.11 12:00
Key point
Four startups use NVIDIA hardware to automate sorting of plastics, recyclables, and textiles, diverting thousands of tons from landfills.
Details
- Four NVIDIA Inception startups—AIRECO, EverestLabs AI, Greyparrot, and reverse.fashion—apply vision AI and robotics to sort mixed waste, plastics, and textiles.
- EverestLabs AI has deployed about 150 robots in North America and Australia, identifying approximately 40 billion objects and recycling roughly 25,000 tons of material.
- AIRECO uses NVIDIA Jetson Orin NX modules for plastic sorting with over 95% recognition accuracy, aiming to increase processing capacity from 18,000 to 48,000 tons annually.
- Greyparrot’s Analyzer systems have made over a trillion bounding box detections across more than 300 units, reducing waste inspection time from 375 hours to six for 15 tons.
- reverse.fashion automates textile sorting at 3 meters per second, processing in an hour what previously took a week, using NVIDIA TensorRT and Jetson modules.
- The technologies address the global challenge where less than 7% of materials entering the economy come from recycled or reused sources.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
NVIDIA LPR Introduces VFlash: A Generalist Decoding Framework for Physical AI Visual PlanningRunway Releases Praxis-1 Open-Weight Robotics Model