ARIS: Automated Recycling Identification System for E-Waste Classification using Deep Learning

Traditional electronic recycling processes suffer from high resource losses due to insufficient material separation and identification capabilities, which limit material recovery. Introducing ARIS (Automated Recycling Identification System), a low-cost, portable e-waste filter that addresses this efficiency gap. The system uses the YOLOx model to classify metals, plastics, and circuit boards in real time, achieving low inference latency with high detection accuracy. Pilot testing yielded 90% absolute accuracy, 82.2% mean average accuracy (mAP), and 84% filtering purity. By combining deep learning with standardized filtering methods, ARIS improves the efficiency of material acquisition and lowers barriers to improved reuse. This work is in line with wider initiatives to extend product life cycles, support exchange and recycling initiatives, and reduce environmental impact throughout the supply chain.