Dwi Nurcahyono, Revanza Wildan Putra Wardana, Sandi Setiawan, Soraya Norma Mustika, Achmad Hamdan, Arya Kusumawardana
This research focuses on the application of a flight controller on a Raspberry Pi-based autonomous flying robot prototype designed to automatically validate hotspots. This system develops an Autonomous Aerial Vehicle (AAV) that can operate independently to assist in monitoring forest and land fires. Raspberry Pi serves as the data processing center for the camera, IMU sensor for stabilization, and GPS module for navigation. The control algorithm uses data from the flight controller to manage the flight path and stability in real-time. Tests were conducted to assess control response, navigation accuracy, and the system's ability to detect and reach the mapped hotspots. One of the latest findings in this research is the interaction between the aircraft control and the Raspberry Pi. The aircraft control functions as the AAV navigation controller, and the Raspberry Pi handles the hotspot validation process using the YOLOv8 algorithm. Our assessment indicates that YOLOv8 achieves outstanding accuracy. The model has successfully classified precision with an accuracy of 85%, recall of 66%, and mAP of 67% for the threshold at IoU 0.5. This integration method has not been widely used in previous research, and it is expected to serve as a foundation for the development of smart and effective AAV systems for environmental monitoring. © 2025 IEEE.
State University of Malang, Electronic Syst. Eng. Tech. Study Prog, Malang, Indonesia; State University of Malang, Power Gen. Eng. Tech. Study Prog, Malang, Indonesia
