Aviation Crash Analysis Dashboard

A comprehensive dashboard for analyzing aviation crash data with AI-powered risk prediction and interactive visualizations.

Features

• Data Collection & Integration

  • Multi-source data collection (NTSB, FAA, ASN)
  • Automated data updates
  • MongoDB integration for efficient storage
  • Data standardization and cleaning

• Interactive Dashboard

  • Real-time crash data visualization
  • Advanced filtering options
  • Trend analysis
  • Aircraft type and operator statistics
  • Export functionality for filtered data

• AI-Powered Analysis

  • Risk prediction for incidents
  • Feature importance analysis
  • Confidence scoring
  • Natural language processing for crash descriptions

• Performance Optimizations

  • Multi-level caching system
  • Database indexing
  • Efficient query optimization
  • Background processing for long-running tasks

Installation

1. Clone the repository:

   git clone https://github.com/Mizokuiam/Aviation-Crash-AIAnalysis.git
   cd pycrashai

2. Create and activate a virtual environment:

   python -m venv venv
   venv\Scripts\activate  # Windows
   source venv/bin/activate  # Linux/Mac

3. Install dependencies:

   pip install -r requirements.txt

4. Set up MongoDB:

   • Install MongoDB Community Edition
   • Start MongoDB service
   • Default connection: mongodb://localhost:27017/

Project Structure

pycrashai/
├── data/               # Data storage
│   └── raw/           # Raw collected data
├── models/            # Trained ML models
├── src/
│   ├── analysis/      # Data analysis modules
│   ├── collectors/    # Data collection modules
│   ├── dashboard/     # Dashboard application
│   ├── database/      # Database operations
│   └── models/        # AI model definitions
├── tests/             # Unit tests
└── requirements.txt   # Project dependencies

Usage

1. Start the dashboard:

   python src/dashboard/app.py

2. Access the dashboard at http://127.0.0.1:8050/

3. Available features:

   • View crash statistics and trends
   • Filter data by date, severity, aircraft type, etc.
   • Export filtered data to Excel
   • Analyze risk factors using AI predictions
   • View geographical distribution of incidents

Configuration

• Environment variables can be set in .env:

  MONGODB_URI=mongodb://localhost:27017/
  DATA_DIR=data/raw
  MODEL_DIR=models
  

Performance Features

1. Caching System:

   • Source data cache (5 min TTL)
   • Query results cache (1 min TTL)
   • Statistics cache (5 min TTL)

2. Database Optimization:

   • Indexed fields: date, source, severity, aircraft_type, operator
   • Text search capabilities for descriptions
   • Efficient aggregation pipelines

3. UI Optimizations:

   • Lazy loading of components
   • Pagination for large datasets
   • Background processing for computations

Development

1. Run tests:

   pytest tests/

2. Code style:

   flake8 src/
   black src/

Dependencies

• Core:

  • Python 3.11+
  • MongoDB 4.4+
  • Dash 2.14.1
  • Pandas 2.1.4
  • Scikit-learn 1.3.2

• AI/ML:

  • TensorFlow 2.18.0
  • SHAP 0.44.0
  • Spacy 3.8.3

• Data Processing:

  • NumPy 1.26.2
  • Plotly 5.18.0
  • XlsxWriter 3.1.9

Contributing

1. Fork the repository
2. Create a feature branch
3. Commit your changes
4. Push to the branch
5. Create a Pull Request

License

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

Acknowledgments

• NTSB for providing aviation accident data
• FAA for regulatory information
• Aviation Safety Network for additional data sources