Waste Segregation using IOT and Deep Learning Classification: An Extensive Review

“According to the World Bank, estimates state that by 2050, the world will generate 3.88 billion tonnes (3,880,000,000,000 kilograms!) of waste each year – an alarming increase of 73% from 2020.” [1, 2]. With the recent rise in population, the generation of waste and garbage has significantly increased. Ineffective waste management practices can result in significant environmental impacts like pollution, wildlife impact, climate change, human health hazards like the spread of diseases and respiratory issues, and even aesthetic and tourism impacts resulting in economic costs. This has resulted in the need for better and more efficient waste management practices. Currently, several well-organized waste management systems are functional. However, they require human intervention, which might be a health hazard for sanitation workers. This has drawn the attention of scientists and engineers. With the alarming increase in waste, there has also been a good increase in the studies conducted on waste management and its automation. In this paper, we survey many such researches. We find that many attempts have been made to automate the process of waste segregation and sometimes the entire waste management process. A lot of ideas propose the effective use of IoT for this purpose. These projects proposed using different sensors to detect the dampness in the waste and then segregate them into wet or dry wastes. However, due to hardcoded thresholding, this might be inefficient. To counter that, many other projects ideate using deep learning algorithms along with IoT devices and provide comprehensive evaluations of these algorithms. The literature review indicated that immense attention is paid to this issue, and research has been conducted to mitigate the problems in this domain effectively. Despite that, this domain could still use unique approaches to resolve major open questions. Keywords: Waste management, Waste segregation, Waste sorting, Internet of Things (IoT)-based waste management, Deep learning-based waste management, CNN-based waste classification, Smart bins, Automated waste management, Edge computing in waste management, Image-based waste sorting, Sustainable waste management, Waste detection, Cloud Computing in waste management.