EcoSphere AI: Explainable AI for Climate-Smart Housing

EcoSphere AI is an advanced, end-to-end system designed to evaluate global housing safety risks associated with climate change-induced hazards like floods, heatwaves, and droughts. It provides real-time, actionable insights for urban planners, policymakers, and individuals, contributing to the development of safer, more resilient communities.

This project has been accepted for presentation at the 6th Congress on Intelligent Systems (CIS 2025) and for publication in the SCOPUS-indexed Springer Book Series, 'Lecture Notes in Networks and Systems'.

---

📸 Screenshots

Here's a glimpse of the EcoSphere AI User interface.

Main Interface	Analysis Results

---

✨ Key Features

• Multi-Hazard Risk Assessment: Evaluates housing safety against floods, heatwaves, and droughts on a global scale.
• High-Accuracy Predictions: Utilizes an optimized XGBoost classifier that achieves a high predictive accuracy with an Area Under the Curve (AUC) of 0.92.
• Explainable AI (XAI): Integrates SHAP (SHapley Additive exPlanations) to provide transparent, feature-level explanations for each prediction, enhancing user trust.
• Interactive Web Interface: A modern frontend built with Next.js allows users to get instant risk assessments by searching for an address, using geolocation, or clicking on an interactive map.
• Safe Zone Recommendations: Identifies and suggests the nearest safe locations based on the risk analysis.
• Conversational AI Assistant: Features an integrated chatbot powered by Gemini to provide user-friendly decision support and answer questions.
• Scalable Cloud Architecture: Fully deployed on Google Cloud (Vertex AI, BigQuery, Cloud Run) for real-time, scalable inference and robust data processing.

---

🏗️ System Architecture

The system is designed as a modular, cloud-native pipeline that ensures scalability, transparency, and responsiveness.

1. Data Ingestion & Processing: Geospatial data for flood, heat, and drought risks are acquired from sources like JRC Global Surface Water and MODIS via Google Earth Engine. The data is processed and stored in Google BigQuery.
2. AI/ML Inference: The user submits a location from the Next.js frontend. The request is handled by a FastAPI backend, which sends the feature vector to a managed XGBoost model endpoint on Google Vertex AI for real-time prediction.
3. Explainability & Response: For each prediction, the backend computes SHAP values to explain the contribution of each risk factor. The prediction, explanation, risk scores, and safe zone suggestions are returned to the user interface.

---

💻 Technology Stack

Category	Technologies
Frontend	Next.js, Tailwind CSS, Folium, Leaflet.js
Backend	Python, FastAPI
AI/ML	Scikit-learn, XGBoost, SHAP
Cloud & DevOps	Google Cloud Platform (Vertex AI, BigQuery, Cloud Run, Google Earth Engine), Docker
Database & Storage	Google BigQuery, Google Cloud Storage

---

🚀 Getting Started

Follow these instructions to set up and run the project locally.

Prerequisites

• Python 3.9+
• Node.js and npm
• Google Cloud Platform account and a configured service account JSON file (service-account.json).

Installation & Setup

1. Clone the repository:

   git clone [https://github.com/your-username/climate-smart-housing.git](https://github.com/your-username/climate-smart-housing.git)
   cd climate-smart-housing

2. Setup the Backend:

   # Create and activate a virtual environment
   python -m venv venv
   source venv/bin/activate  # On Windows, use `venv\Scripts\activate`
   
   # Install Python dependencies
   pip install -r requirements.txt

3. Setup the Frontend:

   # Navigate to the frontend directory
   cd climate-smart-frontend
   
   # Install Node.js dependencies
   npm install

4. Environment Variables:

   • Place your Google Cloud service-account.json file in the root directory.
   • Create a .env file if needed for other configuration variables (e.g., API keys).

Running the Application

1. Start the Backend Server:

   • From the root directory, run the FastAPI application.

   uvicorn app:app --reload

2. Start the Frontend Development Server:

   • In a new terminal, navigate to the climate-smart-frontend directory.

   npm run dev

   • Open your browser and navigate to http://localhost:3000.

---

📁 File Structure

Here is an overview of the key files in the project repository:

.
├── climate-smart-frontend/ # Next.js frontend application
├── Ne_10m_data/          # Geospatial data files
├── app.py                # Main FastAPI backend application
├── climate_safe_housing_model.pkl # Pre-trained XGBoost model file
├── package.json          # Node.js project configuration
├── requirements.txt      # Python dependencies
├── safe_zones_landonly.csv # CSV file with safe zone coordinates

---

✍️ Authors

This project was developed by:

• Kowshik Padala
• Rahul Thota
• Teja Sai Sathwik P
• Dhanush B
• Divya. Udayan J*

From the Department of Computer Science and Engineering, Amrita Vishwa Vidyapeetham, Amritapuri, India.

---

📜 License

This project is licensed under the MIT License. See the LICENSE file for more details.