Automated E-Waste Disposal System Using Convolutional Neural Networks for Classification and Sorting

Abstract E-waste management in India faces a lot of challenges due to improper disposal that results in environmental hazards. An automated e-waste disposal system is proposed which uses Convolutional Neural Network (CNN) for classification. The system consists of a Raspberry Pi, a camera module, a servo motor and an ultrasonic sensor. This CNN model classifies images of e-waste items, which the camera takes, in recognizing different classes of e-waste with a very high degree of accuracy. The system automatically sorts items into appropriate bins based on the result of classification for efficient recycling. Machine learning algorithms are implemented in the estimation of the recyclability of materials, which improves sorting and reduces the need for human labor. The system attains a 95% accuracy rate in the classification of ewaste and improves resource recovery, reducing the environmental impact due to inappropriate disposal.