Swachh Bharat Smart Bin: Design and Development of an IoT- and Machine Learning-Based Intelligent Waste Segregation and Monitoring System

The present paper will present an IoT- and ML-based smart-bin system to help with the waste sorting when picking waste in urban areas and track the sanitation promptness. The system consists of one supervisory pipeline that is a combination of fill-level measurements, environmental measurements, event measurements, waste-state measurements, and connection tracking as well as alerting. It was tested on 8000 event records on 30 smart bins in five sizes of urban areas in 60 days. The proposed model yielded a overall classification accuracy of 91.49 percent, a median inference time value of 217.95 ms and a small error in fill-monitoring on all the fill bands that it was tested at. They were discovered to be highly discriminated Biodegradable, Plastic, Metal and Glass and the most difficult residual was the General Waste. Traditional behaviour, too, was consistent with the diverse conditions of the network and inclined towards the prioritisation of the collection, as well as maintenance visibility. This study shows that smarter bins can be used as viable informational hubs of cities to make cities achieve cleaner, faster and more accountable waste-management operations.