Horn-Bill is an autonomous drone system designed to accelerate reforestation by identifying barren patches of land and precisely dropping eco-friendly seed bombs.
Built on AI + Raspberry Pi + Robotics, Horn-Bill combines aerial intelligence with sustainable design to address one of humanity’s most urgent challenges — climate change and deforestation.
- Fixed-wing/Quadcopter design with 1.5 kg payload capacity.
- Supports Autopilot (ArduPilot) for autonomous flight and manual flight modes.
- Configured for stable, long-endurance reforestation missions.
- CNC-cut payload box with detachable Raspberry Pi mount for easy upgrades.
- Servo-lever mechanism for accurate seed release.
- Quick-attach wooden payload bay for flexibility and maintenance.
- Runs on Raspberry Pi with PiCam/USB camera.
- Uses Roboflow vegetation segmentation model to identify:
- Green zones (vegetation).
- Non-vegetation patches (ideal drop zones).
- Wind + altitude compensation: physics-based model predicts drift to ensure seeds land precisely.
- Real-time overlays showing “safe drop zones.”
- Handmade, eco-friendly clay–compost seed balls (~3 cm).
- Contain seeds + nutrients + moisture for protection and germination.
- Weight optimized (~10 g) to reduce air resistance and ensure accurate drops.
- Disperse seeds upon ground impact for maximum coverage.
-
Take-off & Stabilization
- Drone lifts off and enters stabilization mode (locks altitude for accuracy).
-
AI Vision Processing
- Camera feed processed on Raspberry Pi.
- Vegetation model segments land, inverts mask to highlight barren zones.
-
Drop Decision
- Wind speed & direction provided via anemometer input.
- Physics module adjusts drop coordinates.
- If target zone is clear (within radius), system marks it as droppable.
-
Seed Release
- On user confirmation → servo triggers payload release.
- Eco-seed bomb falls, disperses seeds on ground.
Since flying drones isn’t always allowed indoors/competitions, Horn-Bill runs a simulation pipeline on a Mac/RPi:
- Load aerial images/videos → segmentation + drop logic simulated.
- User can adjust wind speed, direction, and altitude via UI sliders.
- Console logs and overlays show drop decisions.
- Plants hundreds of seeds per mission with high accuracy.
- Reduces need for manual reforestation labor.
- Uses sustainable seed bombs → no plastic waste.
- Scalable into swarm drone systems for massive reforestation efforts.
- Aligns with UN SDGs:
- SDG 13: Climate Action
- SDG 15: Life on Land
- SDG 11: Sustainable Cities & Communities
- Hardware: Raspberry Pi, PiCam/USB camera, Servo system, Drone frame (custom-built).
- Software: Python, OpenCV, NumPy, Roboflow API, ArduPilot.
- AI: Roboflow Vegetation Segmentation Model.
- Other Tools: CNC fabrication, 3D printing for payload system.
🎥 Horn-Bill Demonstration Video (link to YouTube demo)
├── src/ # Main codebase │ ├── camera/ # Camera & feed capture modules │ ├── ai/ # AI inference & segmentation │ ├── physics/ # Wind + drift calculation │ ├── payload/ # Servo & seed-drop control │ └── simulation/ # Simulation-only modules ├── docs/ # Project report, posters, diagrams ├── seed_bomb/ # Seed bomb recipe & tests └── README.md # This file