Robotics & UAV Autonomy Engineer
Multirotor autonomy · GPS-denied navigation · Drone swarms · Perception
I build autonomous multirotors — from the airframe and flight-controller tuning up through companion-computer perception and multi-vehicle coordination. Most of my work lives where the control stack meets real hardware: sensor calibration, EKF behaviour, flight testing and the unglamorous business of finding out why an aircraft did something unexpected.
Currently working on GPS-denied navigation — optical flow and visual odometry as a fallback when GNSS degrades or disappears.
| Project | What it is |
|---|---|
| 🔥 FireHawk Mk1 | Autonomous wildfire-spotting quadcopter, built from scratch in Webots — custom PID stabilisation, lawnmower search, colour-based fire detection, ground-plane GPS projection |
| 📡 Drone RF Detection | Passive radar using a WiFi hotspot as illuminator of opportunity — detects a drone that isn't transmitting, via propeller micro-Doppler |
| 🎯 UAV Precision Landing | ArUco-guided descent to centimetre repeatability, including the committed final phase once the marker leaves the camera's view |
| 🛡️ UAV Obstacle Avoidance | Reactive LiDAR + ToF avoidance layered underneath GPS waypoint navigation |
| 🗺️ Indoor GPS-Denied Navigation | RTAB-Map visual SLAM feeding ArduPilot's EKF as an external nav source |
| 👥 Human-Aware Navigation | CoHAN social navigation on a PAL Tiago — planning around people, not just around obstacles |
| 👁️ OAK-D Perception | OAK-D depth + on-camera inference paired with VINS-Fusion visual-inertial odometry |
Also: a 25-drone swarm system flown simultaneously in field tests, with a LoRa ground control station operating the whole fleet over a ~300 B/s link.
Flight stack ArduPilot firmware and parameter configuration · EKF source selection and failsafe behaviour · sensor and ESC calibration · compass/magnetic interference diagnosis · dataflash log analysis · bench and flight testing
Autonomy GPS-denied navigation · optical flow and visual-inertial odometry · visual SLAM (RTAB-Map, VINS-Fusion) · obstacle avoidance · precision landing · multi-vehicle coordination
Perception OpenCV · PyTorch / TensorFlow · semantic segmentation · multi-object tracking · on-device inference (Jetson, OAK-D VPU)
Platforms & protocols Quadcopters 3–17 inch, hexacopters, heavy-lift, FPV · Raspberry Pi 4/5 · NVIDIA Jetson Nano / Orin Nano · MAVLink, pymavlink, MAVROS, DDS · LoRa
Languages & tools Python · MATLAB / Simulink · ROS 1 & ROS 2 · Gazebo · Webots · LaTeX
MSc — Electrical Engineering, AI & Autonomous Systems Mobile Robotics · Advanced Robotics · Computer Vision · Machine Learning
BSc — Electrical Engineering
Achievements — 2nd place, Swarm Competition (AI Hackathon, Islamabad) · TEKNOFEST 2026, Rocket Category