Capella Technology · Autonomous UAV Team

01 / Vehicle

The UAV

Our unmanned aerial system: airframe, avionics, autonomy stack and payload, and the design decisions behind each of them.

Key specifications

Configuration: Fixed-wing VTOL
Propulsion: 3× T-Motor F100 2810 1100KV · HQProp 9050 · 6S3P 15000 mAh Li-ion
Autopilot: Pixhawk Cube Orange Plus
Compute: NVIDIA Jetson Orin NX · computer vision
Payload: SIYI A8 mini gimbal camera · servo-actuated aid drop

Subsystems

Airframe & Structures

A Flightory Stallion-VTOL airframe with an Eppler E205 wing profile for aerodynamic efficiency and low mass; primary parts are 3D-printed in lightweight PLA-LW.

Propulsion & Power

Three T-Motor F100 2810 1100KV motors with Hobbywing Skywalker 60A ESCs and HQProp 9050 propellers, powered by a 6S3P 15000 mAh Molicel P50B Li-ion pack sized for a safe thrust-to-weight margin.

Avionics & Flight Control

A Pixhawk Cube Orange Plus flight controller, Here 3 GPS, a Matek PDB-HEX power board, 12V/25V regulators and a 100A breaker fuse for stable, protected autonomous flight.

Autonomy & Perception

A Jetson Orin NX mission computer running ROS, YOLO11s object detection, BoT-SORT / OSNet ReID tracking and Multi-Objective Model Predictive Control (MOMPC).

Payload & Aid Delivery

A SIYI A8 mini 4K gimbal camera feeds the Jetson detection and tracking pipeline; the resulting target coordinates trigger a servo-actuated, precise aid-drop.

Communications & Ground Control

915 MHz RFD900X telemetry, a 5 GHz Ubiquiti Rocket R5 AC / Nanostation 5AC Lite video downlink, a 2.4 GHz FlySky i6X manual link, and Mission Planner plus a custom interface at the ground station.

Design decisions & trade studies

Decision 01

Fixed-wing vs. VTOL

Fixed-wing platforms offer high aerodynamic efficiency but need a runway or hand launch, a serious operational constraint on the competition field. A VTOL layout adds motor and servo complexity but enables vertical takeoff and safe landing in tight areas; after liftoff the wing motors tilt to horizontal, keeping the high cruise speed (19.8 m/s) and endurance (25.7 min) of a fixed wing.
Decision 02

Autonomy / perception stack

The full perception and decision loop runs on-board (edge) for low latency and link-loss tolerance, so we chose the Jetson Orin NX for real-time deep-learning inference. ROS provides modularity, YOLO11s real-time detection, BoT-SORT / OSNet ReID robust tracking, and MOMPC unifies the multi-objective mission control.

02 / Testing

Validation

CAPELLA's test philosophy runs from component to system and from simulation to field, to minimise crash risk: mechanical and electrical validation on the bench, then simulated data flows, then ground and flight testing before full mission rehearsals.

Test procedures

1
Pre-flight
Weather, mechanical connections, battery mounting, centre of gravity, RC/telemetry link, safety switches and control surfaces are checked from a checklist, with the camera verified before takeoff.
2
Bench / HIL
Motor thrust measurements, wing span-load tests, power regulator outputs, and software/hardware validation against simulated data flows in Gazebo/PX4/SITL.
3
Ground & manual
With propellers off, control-surface response to RC commands is verified; then manual vertical takeoff, hover and position-hold are qualified in QSTABILIZE mode.
4
Autonomous
Switching from QLOITER to FBWA/AUTO, the 90° tilt transition to forward flight, autonomous route following, lock-on and dive scenarios are tested and the flight data reviewed.

Flight media

03 / Blog

Field Notes

Longer write-ups from the team: technical deep-dives, competition recaps and lessons learned along the way.

04 / Team

The Team

A multidisciplinary student team from TOBB University of Economics and Technology, Ankara: autonomy, avionics, mechanics, communications and perception.

Members: 11+
Platform: Autonomous UAV
Stack: Jetson · CV
Mission: Detect · Deliver

Faculty advisor

Assoc. Prof. Dr. Recep Görgülüarslan

Faculty Advisor

Guidance · Mechanical Engineering · Mentorship

Members

Serhat Uğurlu
Team Lead · Autonomy
CV · ROS · Mission
Salim Eren Kaya
Avionics · Power
PCB · EMI · Power
Doğukan Çalışkan
Software · Security
Backend · Cyber Security
Musa Eren Aydoğmuş
Mechanics · Airframe
CAD · CFD · Manufacturing
Mustafa Öztürk
Comms · Telemetry
UART · RF · Link budget
Hamza Boyraz
Perception · AI
Mapping · Tracking · Jetson
Beraat Taş
Mechanics · Airframe
CFD · Manufacturing
Mehmet Emre Şişman
Avionics · Telemetry
Ground control · Tracking · Algorithms
Ali Efe Çavuşculu
Ops · Media
PR · Sponsorship · Content
Metehan Çelik
Intern
Manufacturing
Eyüp Ensar Baş
Intern
Social media · Sponsorship · PR

05 / Mission

Why We Fly

Mission & Vision

Mission

We develop intelligent UAV systems that detect and locate victims and critical assets in disaster scenarios, then deliver the right aid payload safely and precisely. By combining autonomy, avionics, perception and robust system engineering, we turn real-world constraints into reliable mission performance.

Vision

To become a leading student UAV team known for practical autonomy, dependable engineering and meaningful impact, building aerial systems that support search & rescue, humanitarian logistics and time-critical operations.

06 / Media