NVIDIA says AI and robots are improving recyclable waste sorting
🎧 Listen to this story
NVIDIA says computer vision and robotics can identify and sort used clothing, bottles, plastics and other mixed waste, keeping more materials in circulation. The company cites Taiwan-based AIRECO, which uses NVIDIA Jetson Orin NX modules to operate robotic arms; accuracy on a test line exceeds 95%, while the startup plans to nearly triple annual capacity from 18,000 to 48,000 tons and cut operating costs by 85%.
According to NVIDIA, U.S. startup EverestLabs AI uses computer-vision-guided robots at facilities in North America and Australia, with about 150 robots deployed. The company estimates that the systems have identified roughly 40 billion objects, recycled about 25,000 tons of material and avoided approximately 30,000 metric tons of carbon dioxide emissions.
Why it matters · editorial interpretation
The combination of computer vision and robotics could expand waste-sorting capacity and reduce operating costs, according to figures cited by NVIDIA and AIRECO. The estimates also help quantify potential environmental effects, although they were disclosed by the companies themselves.