Design and Implementation Of An IoT-Based Rat Pest Monitoring System With Automatic Image Notification For Archive Rooms: A Case Study Of The Tulungagung Regency Transportation Agency

Abstract

The presence of rats in archive storage rooms can cause damage to important documents, making an automatic monitoring system necessary for early detection. This study aims to design and develop a rat detection prototype based on the Internet of Things (IoT) using the You Only Look Once (YOLO) method. The research employed the Research and Development (R&D) approach, consisting of problem identification, data collection, product design, design validation, design revision, implementation, and product testing. The system utilizes a Logitech C270 HD webcam for image acquisition, a laptop for YOLO-based image processing, an ESP32 microcontroller, a relay, an ultrasonic speaker, Telegram notifications, and a web dashboard. The testing results showed that the system was able to detect rats in real time with an accuracy of 83.33% based on 24 testing scenarios using different viewing angles. The system also successfully sent notifications via Telegram, displayed detection results on the web dashboard, and automatically activated the ultrasonic speaker when a rat was detected. The validation results achieved an average score of 96% from experts and 90% from users, indicating that the prototype is highly valid and feasible for implementation in the archive room of the Tulungagung Regency Transportation Office. Future research is recommended to increase dataset diversity, use higher-resolution cameras, and develop a dedicated mobile application to improve detection accuracy and real-time monitoring.