IoT based System for Early Detection and Monitoring of Mosquito Breeding Sites

Abstract

This research develops an IoT-based dengue prediction and prevention system integrating real-time environmental monitoring, automated species identification, predictive analytics, and community engagement to combat dengue in vulnerable regions. The study addresses critical gaps in existing literature by fusing IoT sensor networks (temperature, humidity, water-level, and MEMS microphones), epidemiological data from the National Dengue Control Unit (NDCU), meteorological records, and community-reported data via a multilingual mobile application. Methodologically, we employed stratified sensor deployment across urban, semi-urban, and rural zones, coupled with machine learning models (SVM/Random Forest for mosquito classification, CNNs for image-based breeding site detection, and time-series forecasting for outbreak prediction). Results demonstrated 89.25% prediction accuracy, dynamic heatmaps for risk visualization, and high community engagement (120 reports from urban areas alone). The system enables data-driven public health interventions, reducing dengue spread through proactive breeding site elimination and real-time risk communication, representing a scalable paradigm shift in vector-borne disease control.