CLEANTRACK: AN ARTIFICIAL INTELLIGENCE AND INTERNET OF THINGS (AIOT)-BASED SMART WASTE MANAGEMENT SYSTEM

Abstract

Effective waste management requires timely monitoring, accurate waste classification, and efficient collection planning. Conventional waste collection methods mainly depend on fixed schedules and manual monitoring, which can result in overflowing bins, unnecessary collection trips, and inefficient use of resources. This paper presents CleanTrack, an Artificial Intelligence and Internet of Things (AIoT)-based smart waste management system designed for real-time waste monitoring and intelligent collection management. The proposed system uses an IoT-enabled smart bin equipped with an ultrasonic sensor for fill-level measurement, a load cell for waste-weight measurement, a temperature sensor for temperature monitoring, and a Global Positioning System (GPS) module for location tracking. The collected sensor data are transmitted for monitoring and analysis, enabling assessment of bin conditions and waste-generation patterns. For waste identification, a YOLOv8 Nano (YOLOv8n) classification model is integrated for image-based waste classification and achieved a Top-1 accuracy of 95.2% on the validation dataset. A Genetic Algorithm is further employed to optimize waste collection routes based on bin conditions and collection requirements. By combining IoT sensing, AI-based waste classification, data analysis, and route optimization, CleanTrack provides an integrated approach for improving the efficiency, responsiveness, and data-driven planning of waste collection operations.