-
Champion — SLIIT ROBOFEST 2025 Autonomous Drone Challenge (Sri Lanka)
Built a fully custom quadcopter with GPS-denied, vision-based precision landing.
Onboard stack: Raspberry Pi 4 (vision) + STM32F7 (flight control) with ArduPilot.
Trained a custom YOLO model for alpha landing markers; ~7 FPS real-time on Pi.
Full pipeline from Gazebo sim to stable field missions. -
XFly Pro — Flight Controller for PX4/ArduPilot
Dual high-grade IMUs (ICM-42688P / ICM-45686), DPS310/BMP388, isolation & companion IO. -
XFly Lite — Betaflight/INAV FC
STM32F4, GPS/ESC/RX ready for DIY multirotors & fixed-wing.
-
Silver Medal — International Invention & Innovation (INNOPA) Expo, Indonesia
Smart helmet for accident detection and automatic emergency alerts. -
Silver Medal — National Invention & Innovation Competition (Sri Lanka Inventors Commission)
Smart helmet: real-time accident response & safety. -
Top 5 — United Nations Green Innovation Startup Awards (Sri Lanka, UNDP)
Agri-focused IoT innovation; secured LKR 0.8M seed funding. -
Top 8 — Plastic Innovation Challenge (European Union)
Sustainable plastic alternatives & circular economy; awarded LKR 1M. -
Champion — SLIIT ROBOFEST 2025 (Autonomous Drone Challenge)
Sri Lanka’s first autonomous drone competition; precision landing with ArduPilot + Raspberry Pi + STM32F7. -
Finalist — UoM Spark Challenge (Tea-Sense)
Computer-vision system for tea-leaf quality analysis.
-
Vision-Based Autonomous Drone (GPS-denied landing)
👉 https://github.com/KiranGunathilaka/autonomous_drone_robofest -
XFly Pro — UAV Flight Controller
👉 https://github.com/mugesram/UAV-FC -
XFly Lite — Entry Flight Controller
👉 https://github.com/mugesram/Drone
Quick tour: check the pinned repos on my profile.
Autonomy & Control: ArduPilot, PX4, DroneKit, MAVLink, PID tuning
Perception: YOLO, OpenCV, PyTorch
Sim & Dev: Gazebo, ROS/ROS2, GitHub Actions
Embedded & HW: STM32 (CubeIDE), Jetson Nano, Raspberry Pi, Altium
Backend: Python, FastAPI/Flask, MySQL/Postgres, REST
- Robust vision-only landing under lighting/occlusion changes
- Reinforcement-learnig based drone control
- Sensor fusion + vibration isolation on custom FCs
- Field-reliable pipelines from Gazebo → real flights
| XFly Pro | XFly Lite |
|---|---|
 |
 |
- Email: mailto:mugeshkrish007@gmail.com , mailto:mugesramk.22@uom.lk
- LinkedIn: https://www.linkedin.com/in/Mugesram/
Deans List (Sem 1–4), CGPA 3.94/4.0 — building practical autonomy with real deployments.