Recycling centers face significant inefficiencies due to improper material sorting, which increases waste and slows processing. We designed an automated recycling device to improve sorting accuracy and reduce manual labor in recycling plants.
We built a machine that sorts common recyclable items including shredded cardboard, small soda cans, mini water bottles, and small glass bottles. Items enter through a hopper and travel on a conveyor belt where the system weighs each object and captures an image to estimate its size. We calculate density using weight and size measurements while simultaneously sending images to an artificial intelligence program for material prediction. The system combines density calculations with AI predictions to identify material types, then uses paddles to direct items into designated bins. Each bin measures the total weight of collected materials and alerts workers when full.
Our testing demonstrated that the machine sorted plastic, aluminum, glass, and cardboard with at least 80 percent accuracy. The system also separated unidentifiable items into a separate bin. Results showed that combining AI analysis with density calculations improves sorting performance compared to using either method alone. We designed the system with emphasis on safety, reliability, and ease of operation for recycling facility workers.
This project demonstrates how automation can enhance recycling operations by reducing waste, saving processing time, and improving material recovery rates through more accurate sorting.
