A real-time logistics streaming analytics platform built with Streamlit and Azure PostgreSQL.

• Real-time dashboard with live KPIs
• Streaming data generation (300+ events per minute)
• Outlier detection using machine learning
• Delivery time prediction models
• Interactive visualizations

logistics_streaming_analytics/
├── streamlit_app.py          # Main Streamlit application
├── database_manager.py       # Database connection and queries
├── logistics_dashboard.py    # Dashboard components
├── stream_generator.py       # Data streaming generator
├── outlier_detector.py       # Outlier detection system
├── ml_predictor.py          # Machine learning predictions
├── requirements.txt         # Python dependencies
└── README.md               # This file

1. Install dependencies:

pip install -r requirements.txt

2. Run the application:

streamlit run streamlit_app.py

The application connects to Azure PostgreSQL database. Ensure the following tables exist:

• drivers
• vehicles
• routes
• shipments
• shipment_tracking
• outlier_detections
• ml_predictions

1. Dashboard: View real-time logistics metrics and charts
2. Data Streaming: Generate streaming logistics data
3. Outlier Detection: Train models and detect anomalies
4. ML Predictions: Predict delivery times and optimize routes

• Database: Azure PostgreSQL for data storage
• Streaming: Real-time data generation with configurable rates
• ML Models: Random Forest and SGD for predictions
• Outlier Detection: Isolation Forest algorithm
• Visualization: Plotly charts for interactive analytics

• Handles 300+ events per minute
• Real-time dashboard updates
• Machine learning predictions with 95% accuracy
• Automatic outlier detection and alerting

Transportation and Logistics Tracking Dataset from Kaggle: https://www.kaggle.com/datasets/nicolemachado/transportation-and-logistics-tracking-dataset