top of page
Dr Manjari Chakraborty | Dr Rizwan Kazmi
ISBN:
978-81-180058-1-0
DOI:
18. Edge Machine Learning for Smart Waste Segregation and Emission Control
Circular Economy in the Construction Industry: Waste Management Practices
Rujul Goel | Ajay Pratap Singh | Dr. AmanPreet Kaur
DOI:
2026 | 337 pp.
The rapid nature of changing urban lifestyles, increasing consumerism, and progress in packaging has changed the structure of municipal solid waste to a large extent. Such a change has increased methane emissions due to waste rooms, increasing environmental hazards due to incorrect insulation and uncontrolled organic waste. The current classification systems are subject to delay and are not very accurate in the real-time identification of asymmetrical waste currents. To solve these problems, this study proposes an edge-driven smart system for effective waste classification and methane emission control based on a combination of light machine learning and smart sensing infrastructure. The system uses MobileNetV2, a lightweight convolutional neural network, to sort waste into recyclable and non-recyclable streams using real-time image data from a camera. For sorting wet waste, a source of methane emissions, moisture sensors are used. Identified wet waste is routed to biodegradable bags, minimizing anaerobic decomposition in landfills. Ultrasonic and load sensors are applied to detect fill levels in packing containers, with mechanical sorting only operating when necessary, saving energy. A gas sensor is utilized to track toxic emissions, growing environmental protection and allowing data-driven insights. The complete system is locally managed using a Raspberry Pi board, permitting edge computing and avoiding the use of cloud processing, diminishing latency and the carbon footprint for data transfer. The system achieves a high 98.2% classification rate across six types of waste, which is much higher than current cloud-based deep learning methods. The system facilitates real-time separation of waste, reduces human handling, and encourages environmentally friendly waste disposal. The system offers an expandable, low-cost, and green urban waste management system that contributes to a clean and also ensures an efficient smart city framework.
bottom of page