KNUST Research Week and Scientific Conference: Empowering Researchers to Achieve Impactful Outcomes

Mosquito-borne diseases remain a major global health threat, underscoring the need for effective surveillance tools for species identification and activity monitoring. We present a Mosquito-Tracker which is an IoT enabled, AI-driven bioacoustic and climate monitoring system for real-time mosquito surveillance. The device is built using low-cost hardware suite such as microcontroller, GSM/GPS, microphone, and temperature-humidity sensors. It simultaneously records mosquito wingbeat sounds and environmental data from targeted locations, tagging each recording with GPS coordinates. Data are stored locally but can also be transmitted to a cloud storage platform via Wi-Fi. Mosquito classification is performed using a custom deep learning model trained on wingbeat audio signals. The system classifies mosquitoes by genus, gender, and age group across three primary genera, Aedes, Culex, and Anopheles, as well as a “not-mosquito” class, achieving 92% classification accuracy. By combining wingbeat acoustics with environmental data, the Mosquito-Tracker allows precise surveillance and enables correlation of mosquito activity with environmental conditions. This approach supports early outbreak detection, enhances predictive models of mosquito-borne disease risk, and informs targeted intervention strategies.