An intelligent IoT surveillance system for industrial equipment monitoring, featuring real-time anomaly detection, AI-powered forecasting, and decision support.
Note: This is an English-translated version of the VPI86 Surveillance Dashboard.
IOT surveillance dashboard/
├── backend/ # Flask REST API server
│ ├── server.py # API endpoints (timeseries, anomalies, correlations, forecast)
│ ├── data_loader.py # CSV data loading with multi-header support
│ ├── analytics.py # Z-score anomaly detection, correlations, linear regression forecast
│ ├── timeseries.py # Time series data structures
│ └── dataset/ # → links to shared dataset (data-large.csv)
│
├── gradio-app/ # Gradio interactive dashboard (advanced analytics)
│ ├── app.py # Main Gradio dashboard UI
│ ├── analytics_v3.py # Advanced: LightGBM forecast, Isolation Forest, decision support
│ ├── analytics_v2.py # Intermediate: Holt-Winters forecast, IForest anomaly detection
│ └── dataset/ # → links to shared dataset
│
└── modern-frontend/ # React + Vite + TailwindCSS dashboard
├── src/App.tsx # Main dashboard component
├── index.html # Entry point
└── ... # Config files (vite, tailwind, typescript, eslint)
cd gradio-app
pip install gradio plotly pandas numpy scikit-learn statsmodels lightgbm
python app.py# Terminal 1: Start backend
cd backend
pip install flask pandas numpy
python server.py
# Terminal 2: Start frontend
cd modern-frontend
npm install
npm run dev- Real-time Visualization: Interactive time-series charts with historical data overlay
- AI Forecasting: LightGBM-based predictive analytics with confidence intervals
- Anomaly Detection: Z-score and Isolation Forest methods for detecting sensor anomalies
- Multivariate System Health: Outlier detection across all sensors simultaneously
- Correlation Analysis: Heatmap and ranked correlation insights between sensors
- Decision Support: Automated operational recommendations based on data analysis
- Future Risk Alerts: Early warning system for predicted threshold breaches