[Seminar Review] AI-Powered Pixhawk (PX4) Autonomous Flight Training at Goheung Drone Center

Hello, this is QUAD Drone Laboratory.

On January 22nd (Thursday) ~ 23nd (Friday), Quad Drone Laboratory visited Goheung, Jeollanam-do, the heart of Korea’s aerospace industry. We are excited to share the highlights from our seminar titled “Autonomous Flight of Pixhawk (PX4) Drones using AI: Theory and Practical Application,” held at the Goheung Drone Center within the Jeonnam Technopark (JNTP) Aerospace Industry Center.

Running from 10:00 AM to 5:00 PM, this intensive 6-hour session was designed for drone developers and industry professionals. We moved beyond basic flight principles to cover in-depth topics including system architecture, ROS2 integration, and AI-based mission planning.

---

🚀 Seminar Overview

• Date: January 22nd, 10:00 ~ 17:00 (6 Hours)

• Location: Goheung Drone Center, Jeonnam Technopark Aerospace Industry Center.

• Speaker: Ha Sang-wook (CEO, Quad Drone Laboratory / Former Professor, Seoul Drone Academy).

• Topic: Autonomous Flight of Pixhawk (PX4) Drones using AI – Theory and Application App Development.

---

📍 Visiting the Hub of the Drone Industry: Goheung Drone Center

(Insert panoramic photo or map of Goheung Drone Center here)

The seminar took place at the Goheung Drone Center, a premier facility equipped with specialized industrial complexes, flight test control centers, and a newly constructed 1.2km runway essential for drone and UAM (Urban Air Mobility) testing. conducting technical training in such a state-of-the-art environment was a meaningful experience.

---

📚 Key Curriculum & Session Highlights

The seminar was based on Quad Drone Laboratory’s “2026 Revised Edition (Ver 2.0)” curriculum. We focused on a full-stack approach necessary for real-world commercial drone development, ranging from the internal architecture of PX4 to high-level AI integration.

1. Deep Dive into Pixhawk (PX4)

We analyzed drone operation principles through the lens of Newton’s laws of motion and Control Engineering (PID Control). The session covered the detailed structure of the PX4 Flight Stack and the EKF2 (Extended Kalman Filter) estimator. We also compared ArduPilot and PX4, providing guidelines for selecting the right platform for commercial product development.

2. MAVLink Communication & MAVSDK Programming

Participants learned to analyze the structure of MAVLink, the standard protocol for drone communication. We conducted practical coding sessions using the Python-based MAVSDK library to control the drone programmatically:

• Automated Takeoff and Landing.

• GPS coordinate-based Mission Planning.

• Offboard Control for precise velocity and position manipulation.

3. System Expansion with ROS2 and uORB

We explored the bridge technology between ROS2 (Robot Operating System 2)—the standard for modern robotics—and uORB, the internal messaging bus of PX4. The training covered high-level control architectures involving data exchange between the Flight Controller and Companion Computers using MAVROS and Micro-XRCE-DDS Agents.

4. AI-Based Autonomous Patrol Application

In the final session, we demonstrated the development of a “Park Patrol Drone App.” utilizing YOLO and OpenCV. Participants saw how to implement computer vision to detect and track people or vehicles, showcasing the potential for developing intelligent drone services beyond simple flight .

---

🎓 Conclusion

This seminar with Jeonnam Technopark was a valuable opportunity to share deep technical insights with passionate developers. We would like to thank all the participants who remained engaged throughout the long session.

Quad Drone Laboratory provides professional training courses on Pixhawk, ROS2, and AI drone development for corporations, universities, and research institutes.

From the basics of autonomous flight to advanced AI implementation, Quad Drone Laboratory is your partner in development.

---

📞 Contact & Inquiries

• Website: www.quad-drone-lab.co.kr

• YouTube Channel: Quad Drone Laboratory

• Naver Developer Cafe: Pixhawk Autonomous Drone Developer Group

• Email: maponarooo@naver.com

• Phone: +82-31-680-1311

---