Sistem Otonom Drone Holybro S500 Untuk Pencarian Manusia Berbasis Kamera Monokular Tetap Menggunakan Raspberry Pi 5

Pradipta, Mahija Ibad (2026) Sistem Otonom Drone Holybro S500 Untuk Pencarian Manusia Berbasis Kamera Monokular Tetap Menggunakan Raspberry Pi 5. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5024221026-Undergraduate-Thesis.pdf - Accepted Version Restricted to Repository staff only Download (81MB) | Request a copy

Abstract

Kemajuan teknologi Unmanned Aerial Vehicle (UAV) atau drone memberikan peluang besar untuk mempercepat operasi Search and Rescue (SAR) di area bencana yang sulit dijangkau. Penelitian ini bertujuan merancang dan mengimplementasikan sistem otonom pada drone Holybro S500 untuk mendeteksi manusia secara real-time berbasis edge computing menggunakan Raspberry Pi 5 dan kamera webcam. Sistem mengintegrasikan Flight Controller Pixhawk 6C dan kendali penerbangan via protokol MAVLink melalui pustaka MAVSDK-Python. Berdasarkan evaluasi, model YOLOv8n hasil fine-tuning berformat ONNX dengan resolusi 320x320 piksel dipilih karena lebih konsisten mempertahankan bounding box pada pengujian luar ruangan dibanding model pose estimation. Perangkat lunak sistem dikembangkan menggunakan arsitektur multi-threading asinkron, didukung distribusi daya stabil berbasis regulator UBEC 5A untuk mencegah kegagalan sistem akibat voltage sag. Hasil pengujian terbang nyata memvalidasi kemampuan navigasi Scout Mission secara kontinu melalui siklus state machine yang meliputi tahap pemindaian, pendekatan, dokumentasi, dan perpindahan area tanpa perlu mendarat setiap kali mendokumentasikan target. Selain itu, sistem keselamatan berlapis mampu memitigasi anomali pembacaan sensor akibat barometric drift saat terbang rendah. Penelitian ini mengembangkan purwarupa drone otonom yang mengintegrasikan persepsi visual, pengambilan keputusan otonom, dan transmisi pemantauan real-time untuk mendukung misi pencarian korban
=================================================================================================================================
The advancement of Unmanned Aerial Vehicle (UAV) or drone technology presents significant opportunities to accelerate Search and Rescue (SAR) operations in hard-to-reach disaster areas. This research aims to design and implement an autonomous system on a Holybro S500 drone for real-time human detection based on edge computing using a Raspberry Pi 5 and a webcam. The system integrates a Pixhawk 6C Flight Controller with flight control executed via the MAVLink protocol through the MAVSDK-Python library. Based on evaluations, a fine-tuned YOLOv8n model in ONNX format with 320x320 pixels resolution was selected due to its superior consistency in maintaining bounding boxes during outdoor testing compared to a pose estimation model. The system software is developed using an asynchronous multi-threading architecture, supported by a stable power distribution system based on a 5A UBEC regulator to prevent system failure due to voltage sag. Real-world flight testing demonstrated continuous Scout Mission navigation through a state machine cycle consisting of scanning, approach, documentation, and displacement phases, without needing to land every time a target is documented. Furthermore, a multi-layered safety system helped mitigate sensor reading anomalies due to barometric drift during low-altitude flights. This research develops an autonomous drone
prototype that integrates visual perception, autonomous decision making, and real-time monitoring transmission to support victim search missions

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Drone Otonom, Komputasi Edge, MAVSDK-Python, Raspberry Pi 5, Pencarian dan Penyelamatan, YOLOv8n, Autonomous Drone, Edge Computing, MAVSDK-Python, Raspberry Pi 5, Search and Rescue, YOLOv8n
Subjects:	T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques. Image analysis--Data processing. T Technology > TL Motor vehicles. Aeronautics. Astronautics > TL776 .N67 Quadrotor helicopters--Automatic control U Military Science > UG1242 Drone aircraft--Control systems. (unmanned vehicle)
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Computer Engineering > 90243-(S1) Undergraduate Thesis
Depositing User:	Mahija Ibad Pradipta
Date Deposited:	16 Jul 2026 01:42
Last Modified:	16 Jul 2026 01:42
URI:	http://repository.its.ac.id/id/eprint/135038

Actions (login required)

View Item