Thumbnail Image

Advanced waste classification using machine learning and environmental impact analysis of MSW disposal methods

Ouedraogo, Angelika Sita
Disposal of municipal solid wastes (MSW) remains a challenge to minimize its impact on the environment and human health. Landfilling, currently the most common method used for MSW disposal, occupies land space, and leads to soil pollution and air emissions. In addition, waste disposal was associated with cancers, respiratory diseases, mortality and morbidity, mental health, and birth anomalies. For human and environmental safety, it is imperative to develop safe waste management practices. The overall goal of this project is to limit landfilling activities in the US by developing advanced algorithms for MSW and plastic sorting. In addition, the study aims to analyze the effect of waste disposal methods on human health and the environment safety and propose sustainable and environmental friendly approaches for waste management in the US. The work was divided into six chapters. Chapter I and VI are the general introduction and conclusions, respectively. Chapter II and III focused on the life cycle assessment of landfilling, gasification, the US conventional waste disposal methods and integrated waste management methods for MSW disposal. Chapter IV and V consisted of developing convolutional neural network (CNN) algorithms for MSW and plastic waste sorting, respectively. The project will help city authorities, waste management planners and environmentalists in developing rules and regulations for waste management in the US. In addition, the CNN algorithms will help in the implementation of fully automated recycling plants. The work is expected to provide a detailed report on the impact of waste disposal methods on human health and the environment. Two algorithms for MSW, and plastic segregation will be provided.