Design and implementation of an internet of things-based automatic waste sorting system

Abstract

This paper presents the design and development of an internet of things (IoT)-based automatic waste sorting system that classifies waste into four categories: organic, non-organic, metal, and others. The system integrates an Arduino Mega for control, multiple proximity sensors (inductive, capacitive, and infrared), and ultrasonic sensors for level detection, and a NodeMCU ESP8266 for real-time monitoring via the Blynk platform. A total of 100 tests (25 per bin) were conducted. Classification success rates were 92% (metal), 80% (inorganic), 84% (organic), and 100% (others), resulting in an overall accuracy of 89%. The main contribution is a combined automatic sorting and IoT monitoring framework suitable for campus-scale deployment.