Godfrey Perfectson Oise and Susan Konyeha
Electronic waste (e-waste) poses significant environmental and health risks due to improper disposal and recycling methods. This research addresses these challenges by developing an intelligent, automated sorting system using a deep learning algorithm, specifically a Sequential Neural Network (SNN) implemented with TensorFlow and Keras. The system is designed to classify various e-waste components, such as circuit boards, batteries, and mobile devices, based on a comprehensive dataset comprising over 3000 e-waste images. The proposed SNN architecture integrates convolutional layers for feature extraction, followed by pooling and dropout layers, leading to fully connected layers and a softmax output for multi-class classification. The research details the design and implementation phases of the SNN, emphasizing its potential to significantly improve the efficiency of e-waste sorting and promote environmental sustainability. Key performance metrics include an overall accuracy of 87%, precision of 87%, recall of 86%, and an F1-score of 86%. These results highlight the model's effectiveness in accurately classifying e-waste categories. This technologically advanced approach aims to revolutionize e-waste management by providing a long-term, cost-effective solution. By facilitating intelligent collection, segregation, and disposal of e-waste, the system fosters a cleaner, safer, and greener environment. The research underscores the importance of integrating advanced machine learning techniques into waste management practices to address pressing environmental issues and promote sustainable development.