Logo Lanfrica

Kimkanja/SlideGuard

Domain:

environment and energygeospatial

Record type:

software
Creator:
Kim
Host:
SlideGuard – Landslide Early Warning System using satellite data, AI, and SMS alerts for high-risk zones in Kenya # SlideGuard – Landslide Early Warning System SlideGuard is a technical solution designed to predict and prevent landslide disasters in Kenya, particularly in high-risk areas like Murang'a and Elgeyo Marakwet. Heavy rains often trigger deadly landslides, burying houses and endangering communities. SlideGuard provides early warnings using satellite data, AI, and automated alerts. ## Features **Satellite Inputs:** - **Sentinel-1 SAR:** Detects soil moisture even through cloud cover. - **Copernicus DEM:** Provides slope maps to identify high-risk areas (>30° slope). **Processing Workflow:** 1. Calculate **Soil Moisture Index** from Sentinel-1 data. 2. Overlay the soil moisture data on slope maps to detect danger zones. **AI Integration:** - Logistic Regression model triggers alerts when: - Soil Moisture > 80% - Slope > 30° **Output:** - Automated SMS alerts for local chiefs in high-risk zones. ## Why SlideGuard Wins - Landslides are a recurring, deadly problem in Kenya. - Uses advanced radar satellite data (Sentinel-1 SAR) instead of basic imagery. - Provides real-time alerts for preventive action. - Combines technical rigor with a practical, community-focused solution. ## Tech Stack - Python (data processing & AI model) - Logistic Regression for risk prediction - Satellite data integration (Sentinel-1 SAR & Copernicus DEM) - SMS alert system (Twilio) ## Installation Follow these steps to set up SlideGuard locally: 1. **Clone the repository:** ```bash git clone github.com cd SlideGuard 2. Create a virtual environment (recommended): python -m venv venv source venv/bin/activate # On Windows: venv\Scripts\activate Install dependencies: pip install -r requirements.txt Make sure your requirements.txt includes all necessary libraries like numpy, pandas, scikit-learn, rasterio, twilio, flask (if using a web interface), etc. Set up environment variables (for SMS API credentials): export TWILIO_ACCOUNT_SID="you …