Browse categories
Explore
Fiverr Pro
English
$
USD
I will integrate computer vision and edge ai for autonomous drones
Autonomous Robotics and UAV Research Engineer
I am a Robotics & UAV Research Engineer and NUST High Achievers Gold Medalist specializing in autonomous systems, Computer Vision, and industrial control. As a TEKNOFEST Finalist with a patented robot...
Level 1
Has met certain performance criteria and shows strong potential in the marketplace.
About this Gig
Autonomous UAV Computer Vision & Edge AI Engineering
Most drones follow GPS, but mine can "see."
I am a Robotics Research Engineer, a Mechanical Engineering graduate from NUST, and a two-time international TEKNOFEST UAV Finalist. I specialize in developing resilient autonomous architectures that bridge the gap between heavy edge computing and physical flight controllers.
My AI & Vision Stack:
- Hardware: NVIDIA Jetson Nano, Raspberry Pi, OAK-D, Coral TPU
- Frameworks: OpenCV, YOLO (v8-v11), MediaPipe, TensorFlow Lite
- UAV Interface: MAVLink, ROS 2, and custom Python API bridges
What I Can Build For You:
- Precision Landing: Detection of ArUco markers or custom pads for accurate landing.
- Object Tracking: Real-time following of people, vehicles, or other drones using Deep Learning.
- Hand Gesture Control: Intuitive swarm or single drone control based on human hand trajectories.
- Dynamic Interception: High-speed detection and path planning for interception missions.
Why choose me? You are hiring a researcher who has built and flown these systems in international competitions and holds a design patent for robotic mechanisms.
Please message me to discuss your hardware setup before ordering!
Platform:
Raspberry Pi
Sensors:
Accelerometer
•
Camera
•
Position
•
Location
•
Gyroscope
Expertise:
Image processing
•
AI
•
Robotics
My Portfolio
FAQ
Which companion computer should I use for real-time AI?
For high-speed object detection like YOLO, I strongly recommend the NVIDIA Jetson series due to their dedicated CUDA cores. However, I can also optimize lighter OpenCV and MediaPipe scripts for the Raspberry Pi if your mission allows for lower frame rates.
Can you help me train a custom YOLO model for my specific target?
Yes. If you have a specific target (e.g., a unique landing pad or a specific type of vehicle), I can train a custom YOLO model using your dataset and optimize it for edge deployment to ensure maximum inference speed on your drone.
How does the AI communicate with my Pixhawk flight controller?
I use MAVLink (via pymavlink, DroneKit or MAVSDK) to send "Offboard" velocity and position commands. The AI processes the camera feed on your companion computer and tells the flight controller exactly how to move to track or intercept the target.
Do I need to send you my physical drone for testing?
No. As an engineer with experience in Gazebo Harmonic and Webots, I develop and rigorously test computer vision and path-planning logic in high-fidelity 3D simulations first. Once the logic is verified, I provide the deployment scripts and a detailed integration guide for your hardware.
Can you implement gesture-based swarm control?
Absolutely. Drawing from my research in human-swarm interaction, I can integrate MediaPipe to allow a single drone or a swarm (like DJI Tello) to mirror your hand trajectories or follow specific gestural commands in real-time.
