Dean Machines: Autonomous FPV Drone Platform

Open-Source Autonomous Drone Platform

🚁 Project Overview

Dean Machines is developing an open-source autonomous drone platform using a 5" FPV Racing drone equipped with advanced sensors and AI capabilities. Our mission is to create a comprehensive dataset for drone autonomy research.

Hardware Stack

• Drone Platform: 5" FPV Racing Drone
• AI Computer: NVIDIA Jetson Orin Nano
• Sensors:
  • TFmini-S LiDAR
  • AI-enabled Camera Module
  • IMU/Gyroscope
  • GPS Module
• Communication:
  • YHY 9800 Eng D SDR
  • 433MHz Receiver
  • FPV Video Receiver
  • 30ft Monopole Antenna

Data Collection Framework

We're building a standardized dataset including:

• Visual Data (RGB + Depth)
• LiDAR Point Clouds
• Radio Telemetry
• Flight Controller Data
• Environmental Metrics

🛠️ Technical Architecture

Onboard Systems

• NVIDIA Orin Nano for real-time processing
• Custom sensor fusion pipeline
• Edge AI inference
• Real-time telemetry streaming

Ground Station

• SDR signal processing
• Real-time data visualization
• Neural network training
• Dataset validation tools

📊 Dataset Contribution

Requirements

1. Hardware Compatibility
2. Sensor Calibration
3. Data Format Standards
4. Quality Metrics
5. Validation Process

📊 Dataset Contribution Guidelines

Required Hardware Configuration

• 5" FPV Racing Drone Frame
• NVIDIA Jetson Orin Nano
• TFmini-S LiDAR Sensor
• HD FPV Camera (>720p)
• GPS Module (uBlox NEO-M8N or better)
• IMU (MPU6050 or better)
• SDR (YHY 9800 or compatible)

Data Collection Requirements

interface DroneDataPoint {
  timestamp: number;        // Unix timestamp (ms)
  gps: {
    lat: number;           // Latitude
    lon: number;           // Longitude
    alt: number;           // Altitude (m)
    accuracy: number;      // GPS accuracy (m)
  };
  imu: {
    acceleration: Vec3;    // m/s²
    gyroscope: Vec3;      // rad/s
    magnetometer: Vec3;    // μT
  };
  lidar: {
    distance: number;      // Distance in meters
    strength: number;      // Signal strength
  };
  camera: {
    resolution: string;    // "1280x720"
    fps: number;          // Frames per second
    format: string;       // "h264"
  };
  radio: {
    frequency: number;    // MHz
    signalStrength: number; // dBm
    bandwidth: number;    // MHz
  };
}

Data Submission Process

1. Data Collection

   • Minimum 10 minutes of continuous flight
   • Various flight patterns required:
     • Hover
     • Forward flight
     • Figure-8
     • Obstacle navigation

2. Data Validation

   # Validate dataset structure
   npm run validate-dataset path/to/data
   
   # Generate validation report
   npm run generate-report

3. Submission Format

   dataset/
   ├── metadata.json       # Flight information
   ├── raw/               # Raw sensor data
   │   ├── camera/        # Video streams
   │   ├── lidar/         # LiDAR point clouds
   │   ├── imu/           # IMU readings
   │   └── radio/         # SDR captures
   └── processed/         # Processed data
       ├── trajectory/    # Flight path
       ├── obstacles/     # Detected obstacles
       └── annotations/   # Manual annotations
   

Quality Standards

• Camera: 720p minimum at 30fps
• LiDAR: 100Hz minimum sampling rate
• IMU: 200Hz minimum sampling rate
• GPS: 10Hz minimum update rate
• Radio: 433MHz band captures at 2MSPS

Review Process

1. Automated validation check
2. Data quality assessment
3. Manual review by core team
4. Integration into main dataset

View Full Documentation on our GitHub Wiki

Components

This project includes the following components:

• SensorDataDisplay: Displays sensor data in a consistent format, and supports units, different display types, and text sizes.
• DroneMap: Displays a map with the drone's location, and supports displaying a path of previous locations and different map types.
• TelemetryDisplay: Displays telemetry data and supports a custom color for the value text.
• LidarDataDisplay: Displays lidar data as a 3D scatter plot with a dropdown menu to select different color scales.
• IMUDataDisplay: Displays IMU data as a 3D scatter plot with a dropdown menu to select different data types to display.
• FPVVideoDisplay: Displays a video stream from the FPV drone, with a loading state, error handling, and a responsive video element. The FPV page also includes an input box to allow the user to enter a video URL and save it for later use, and checkboxes for advanced options for receiving and displaying FPV video in real time.

🔧 Development

Documentation Guides Contact

# Website Development (Next.js)
npm run dev

# Python CV Pipeline (Coming Soon)
python3 setup.py develop

🤝 Contributing

We welcome contributions to Dean Machines! Please follow these steps:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

Read our Contributing Guidelines for detailed information.

Code of Conduct

Please read our Code of Conduct to keep our community approachable and respectable.

📝 License

This project is licensed under the MIT License - see the LICENSE file for details

View Full Requirements