Flagship AI System

🌍 Agroforestry Smart Decision Support System

AI for climate-smart agriculture and sustainable land use in Kenya

ASDSS

📌 Overview

This project transforms machine learning research into a real-world decision support system that recommends optimal tree species for agroforestry systems. It integrates remote sensing, climate data, and terrain features to support sustainable farming and climate resilience.

🚀 Key Capabilities

  • Location-based tree species recommendation
  • Integration of satellite (Sentinel) and environmental data
  • Species distribution modeling using ML
  • Interactive web-based deployment

⚙️ Methodology

  1. Extract environmental predictors (climate, elevation, indices)
  2. Train ML models (RF, SVM, GBM, MaxEnt)
  3. Generate habitat suitability maps
  4. Deploy via Shiny application

📊 Results

  • Random Forest achieved AUC > 0.99
  • High recall and specificity
  • Reliable ecological predictions

🛠 Tech Stack

Python, R (Shiny & Plumber API), Google Earth Engine, Remote Sensing, Machine Learning