Objective:
ForestShield aims to provide an automated, real-time, and scientifically robust platform for monitoring deforestation using satellite imagery and advanced machine learning. By leveraging cloud-native architecture and unsupervised learning, ForestShield delivers accurate, scalable, and actionable insights to help protect global forests.

ForestShield is a fully serverless, cloud-native platform for automated monitoring and analysis of deforestation using Sentinel-2 satellite imagery. It leverages AWS managed services, machine learning (K-means clustering), and scalable event-driven architecture to provide real-time forest monitoring capabilities.

• 🌐 Live Application: https://forestshieldapp.com
• 🚀 API Docs on Swagger UI: https://api.forestshieldapp.com
• 📂 Frontend GitHub Repository: https://github.com/youneslaaroussi/forestshield-frontend

Before setting up ForestShield, ensure you have the following installed and configured:

• AWS CLI (v2.x) - Installation Guide
• Docker (v20.x+) - Installation Guide
• Node.js (v18.x+) - Installation Guide
• Python (v3.9+) - Installation Guide
• Java (v11+) - Installation Guide
• Maven (v3.6+) - Installation Guide

• AWS Account with appropriate permissions
• AWS CLI configured with credentials:

  aws configure
  # Enter your AWS Access Key ID, Secret Access Key, and default region

• Docker authenticated with ECR:

  aws ecr get-login-password --region us-west-2 | docker login --username AWS --password-stdin <account-id>.dkr.ecr.us-west-2.amazonaws.com

Your AWS user/role needs permissions for:

• CloudFormation (full access)
• Lambda, Step Functions, SageMaker
• S3, DynamoDB, Athena, Glue
• App Runner, ECR, ElastiCache
• IAM (for role creation)
• SNS, CloudWatch

git clone https://github.com/youneslaaroussi/forestshield-server.git
cd forestshield-server

# Copy and edit the sample environment file
cp env.sample .env

# Edit .env with your AWS credentials and configuration
nano .env

# Deploy on AWS (This Script will perform all required setup and uses CloudFormation)
./deploy.sh

# if on Windows Powershell and have WSL installed
bash deploy.sh

curl -X POST https://api.forestshieldapp.com/sentinel/step-functions/trigger \
-H "Content-Type: application/json" \
-d '{
  "searchParams": {
    "latitude": -6.0,
    "longitude": -53.0,
    "startDate": "2024-01-01",
    "endDate": "2024-01-15",
    "cloudCover": 20
  },
  "maxImages": 3
}'

View the Step Functions workflow live in your AWS Dashboard.

curl "https://api.forestshieldapp.com/dashboard/heatmap?north=-5.9&south=-6.1&east=-52.9&west=-53.1&days=30"

# Get recent alerts
curl "https://api.forestshieldapp.com/dashboard/alerts?limit=10"

# Get high-priority alerts
curl "https://api.forestshieldapp.com/dashboard/alerts?level=HIGH&acknowledged=false"

For complete API documentation see DOCS.md.

ForestShield represents a state-of-the-art serverless architecture that leverages the full power of AWS managed services to create an intelligent, scalable, and cost-effective deforestation monitoring system. The platform demonstrates advanced cloud-native design patterns and scientific computing at scale.

• 100% Serverless: Zero server management, automatic scaling, pay-per-use pricing
• Event-Driven: Asynchronous processing with decoupled microservices
• ML-First: Machine learning integrated throughout the data pipeline
• Real-Time: Live monitoring with WebSocket connections and instant alerts
• Multi-Region: Geographically aware processing with region-specific ML models

The system implements sophisticated unsupervised machine learning using K-means clustering on scientifically-derived vegetation indices:

NDVI = \frac{(NIR - Red)}{(NIR + Red)}

• NIR: Near-Infrared Band (Sentinel-2 B08, 842nm)
• Red: Red Band (Sentinel-2 B04, 665nm)
• Range: -1.0 to +1.0 (vegetation typically > 0.2)

Each pixel becomes a feature vector: [NDVI, Red_Reflectance, NIR_Reflectance, Latitude, Longitude]

# Sum of Squared Errors optimization
SSE = Σ(k=1 to K) Σ(x∈Cluster_k) ||x - centroid_k||²

At its core, ForestShield is powered by AWS Lambda — handling everything from STAC-based satellite image searches to NDVI computations and unsupervised ML.

Our Java-based image ingestion Lambda is accelerated using AWS Lambda SnapStart, cutting cold starts from 2s to ~200ms. This ensures we can query, fetch, and process fresh Sentinel-2 imagery in seconds.

Thanks to this architecture, ForestShield:

• Processes over 1M pixels/month with under $10 Lambda spend
• Delivers 10x faster clustering vs EC2-based approaches
• Instantly scales from zero to thousands of parallel detections

🏆 This architecture demonstrates AWS best practices for:

• Serverless-first design with event-driven processing
• ML/AI integration with scientific computing
• Real-time data processing and visualization
• Cost-optimized scalable infrastructure
• Enterprise security and compliance patterns

To remove all AWS resources and avoid ongoing charges:

./deploy/cleanup.sh

⚠️ Warning: This will permanently delete all data, models, and infrastructure.

• 📋 Complete Technical Specification: DOCS.md
• 🏗️ Infrastructure Details: cloudformation.yaml
• 🚀 Deployment Scripts: deploy/

1. Fork the repository
2. Create your feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

This project is licensed under the MIT License - see the LICENSE file for details.

• ESA Sentinel-2 for satellite imagery
• AWS for cloud infrastructure
• Open source community for tools and libraries

Methods applied based on research:

• Detecting deforestation in the Amazon rainforest using unsupervised K-means clustering on satellite imagery
• Md Jelas I, Zulkifley MA, Abdullah M and Spraggon M (2024) Deforestation detection using deep learning-based semantic segmentation techniques: a systematic review. Front. For. Glob. Change 7:1300060. doi: 10.3389/ffgc.2024.1300060
• aws-smsl-geospatial-analysis-deforestation