DOAB (दोआब) — Dual-Climate Geospatial & Civic Safety Platform

Detect → Order → Alert → Back-check

Live Production URL: https://doab.sahnirmaan.live (Direct CloudFront Edge: d38z2t6y9aj34z.cloudfront.net)
Developed by Team Contrarians for WeMakeDevs Hackathon.

---

📸 Platform Interface & Live Visuals

Figure 1: Hardware-accelerated 3D Terrain & Real-Time Inundation Risk Vector Engine

🌐 Volumetric Hazard Clusters	📈 Dynamic Elevation Matrix
Real-time municipal telemetry with 15cm risk thresholds	Copernicus DEM ridgeline elevation analysis & bypass routes

🗺️ 2D Low-Bandwidth Municipal Map	🛰️ 3D Perspective Orbital Tilt
Ultra-low latency raster engine for field responders	Orbital camera angle with live digital elevation terrain

---

🌊 Overview & Vision

DOAB (दोआब) is an AI-powered municipal intelligence and citizen life-safety platform built to tackle India's extreme dual-climate crises: flash floods / urban waterlogging during monsoon and acute drought / tanker logistics deficits during dry seasons.

DOAB ingests ground-truth municipal datasets (6,000+ real civic incident reports across BBMP wards), Open-Meteo ERA5 precipitation data, and Copernicus digital elevation models (DEM) to predict waterlogging risks 60 minutes in advance, calculate high-elevation bypass evacuation corridors, and deliver real-time risk intelligence.

---

☁️ Complete Amazon Web Services (AWS) Architecture

Figure 2: Production-Grade AWS Disaster Management Architecture Blueprint

DOAB is engineered on a resilient, high-availability AWS Cloud infrastructure designed for mission-critical disaster management:

                          [ User / Field Responder ]
                                      │
                                      ▼
                        ┌───────────────────────────┐
                        │    AWS CloudFront (CDN)   │  Global edge caching & SSL
                        │    Distribution ID        │  Zero-jerk route loading
                        └─────────────┬─────────────┘
                                      │
                      ┌───────────────┴───────────────┐
                      ▼                               ▼
        ┌───────────────────────────┐   ┌───────────────────────────┐
        │       AWS S3 Bucket       │   │    AWS EC2 (Ubuntu 24)    │
        │   Static Frontend Build   │   │   Dockerized Services     │
        │   (React 19 + PWA Assets) │   │   FastAPI AI + Node API   │
        └───────────────────────────┘   └─────────────┬─────────────┘
                                                      │
                      ┌───────────────────────────────┴───────────────────────────────┐
                      ▼                                                               ▼
        ┌───────────────────────────┐                                   ┌───────────────────────────┐
        │   AWS RDS PostgreSQL      │                                   │  Amazon Location Service  │
        │   (Engine v18.3)          │                                   │  (Places & Routes Engine) │
        │   Persistent Civic State  │                                   │  Pan-India Geocoding      │
        └───────────────────────────┘                                   └───────────────────────────┘

1. AWS Core Infrastructure

• Amazon CloudFront (CDN): High-speed global content delivery edge with TLS 1.3 termination, HTTP/2 multiplexing, and instant cache invalidation (E20R65BKZ9U404).
• Amazon S3: High-durability origin bucket (s3://doab-frontend-970909364340) hosting single-page application bundles, WebP maps, and PWA offline manifests.
• Amazon EC2 (t3/t4 Ubuntu 24.04 LTS): Hosts production dockerized container meshes running our FastAPI prediction service, geospatial routing algorithms, and proxy gateways.
• Amazon RDS (PostgreSQL 18.3): Enterprise managed database hosting civic telemetry logs, emergency dispatch orders, and verified road accessibility states.

2. High-Value AWS Add-Ons & Life-Safety SOS Mesh

• Amazon Location Service (ALS):
  • Pan-India Reverse Geocoding & Search: Live landmark, address, and municipality search across all Indian states with zero third-party tracking.
  • Civic Place Index: High-precision boundary search for municipal wards, storm drains, and vulnerable underpasses.
• Amazon SNS (Simple Notification Service) — Immediate Broadcast Layer:
  • Broadcasts SOS payloads instantly via emergency SMS to Disaster Response Teams (NDRF/SDRF) and family contacts.
  • Delivers high-priority multi-platform push notifications to emergency crew field apps with zero cold-start delay.
• Amazon SQS (Simple Queue Service) — Zero-Loss Distress Buffer:
  • Buffers and queues thousands of concurrent distress beacons during mass-casualty flash floods or dam releases so not a single SOS alert is ever dropped or lost.
• AWS Lambda (Serverless Emergency Compute):
  • Serverless micro-function triggered instantaneously upon distress beacon ingestion, executing coordinate extraction and nearest safe haven routing independently of primary server load.
• Amazon Cognito Identity Pools: Unauthenticated and role-based IAM temporary credential vending for client-side Amazon Location Service access without exposing static API keys (ap-southeast-2:ff8ccdf5-4e7a-45c8-861d-51221d0ea28d).
• AWS CloudWatch: Real-time telemetry monitoring, health probe alarms, and latency tracking across the EC2 prediction containers.

---

⚡ Key Highlights & Core Capabilities

1. Dual-Engine Geospatial Mapping:

   • 2D Standard Engine (Leaflet): Ultra-low battery consumption, raster tile rendering, and instant responsiveness on low-bandwidth field mobile devices.
   • 3D Vector GPU Engine (MapLibre GL): Hardware-accelerated 3D terrain tilt, orbital angles, and real-time volumetric hazard clusters.

2. DOAB AI Inundation & Hydrological Risk Prediction Engine:

   • 1-Hour Forward Inundation Modeling: Fuses Copernicus 201m DEM topography, soil runoff absorption, and sub-level sump passage geometry.
   • Dual-State Switcher: Instant transition between Monsoon Inundation Mode and Scarcity Drought Mode.
   • Segmented Forecast Pill Tabs: Zero-layout-shift switcher toggling between [DOAB AI INUNDATION] and [IMD LIVE & 7D NWP].

3. India Meteorological Department (IMD) Live Telemetry & Doppler Radar:

   • Ingests official IMD Automatic Weather Station (AWS) observations, 24h accumulated rainfall, min/max surface temperatures, and Doppler Radar / Satellite Nowcasts.
   • 7-Day Synoptic NWP Drawer: Day-by-day numerical weather prediction outlook embedded inside the live forecast dock.

4. Multi-Zone Dynamic Resizable HUDs:

   • 3-zone edge resizing (East rail for width, South rail for height, South-East corner for diagonal scaling) across both Live Risk Monitor and Create Safe Elevation Path panels, preserving map panning gestures.

5. Safe Elevation Path & Bypass Routing:

   • Multi-modal evacuation routing that weights elevation gain and avoids historical inundation points.
   • Interactive waypoint creation (Point A Origin and Point B Destination) directly on the live map or via address autocomplete.
   • Offline Path Bookmarking: Save emergency bypass paths to local device storage for guaranteed offline access during storm network blackouts.

6. Dev Admin Disaster Condition Matrix & Rule Simulator (/admin):

   • Standardized test suites for all 6 extreme hazard events: Cloudburst Inundation, Dam Overflow, Sump Submersion, Drought Logistics, Heatwave, and Water Contamination.
   • 1-Click Broadcast: Dispatches verified telemetry beacons across all connected desktop browsers, mobile PWAs, and native Android APK devices via pub/sub.
   • Interactive Rule Engine Simulator: Real-time sliders for rainfall (mm/h), inundation depth (cm), temperature (°C), and water table drop (m).

7. Ground-Truth Arrival & Model Calibration:

   • "Reached Destination" Workflow: Field users confirm arrival and rate model safety accuracy (Accurate, Slight Water, Flooded/Blocked).
   • Generates local telemetry verification loops that improve model precision over time.

8. EXIF Hardware Metadata Inspector (Route Back-Check):

   • Direct binary parsing of JPEG APP1/EXIF headers upon photo upload.
   • Extracts hardware GPS coordinates, shutter timestamps, and camera make/model for cryptographic ground-truth verification with fallback to privacy-preserving mode.

9. Live Remote Admin Telemetry & Cross-Device Sync:

   • Web Dev Admin Console (/admin on laptop) broadcasts live emergency beacons, triggering native haptic alarms, high-priority phone notifications, and full-screen SOS takeovers on field Android devices in < 200ms.

10. Voice AI Operations Copilot:

    • Speech-synthesized civic operations assistant capable of querying nearby ward telemetry, reading route risks aloud, and guiding emergency evacuations.

11. Native Android APK (Capacitor + Google Play Services):

    • Native debug APK with offline asset caching, background geolocation, native Google OAuth bottom-sheet login, and full SOS haptics.

12. Benchmarking & Competitive Analysis:

    • Full competitive whitepaper benchmarking DOAB against Google Flood Hub, NDMA Sachet, IMD Mausam, CWC India-WRIS, and Tomorrow.io:
      • Markdown Report: docs/COMPETITIVE_ANALYSIS.md
      • Word Document: docs/COMPETITIVE_ANALYSIS.docx
      • System Flowchart: docs/COMPETITOR_LANDSCAPE_FLOWCHART.jpg

---

🛠️ Complete Tech Stack

Layer	Technologies
Frontend	React 19, TypeScript, Vite 8, Tailwind CSS v4, Framer Motion
Cloud & Hosting	AWS CloudFront, AWS S3, AWS EC2 (Ubuntu 24), AWS RDS PostgreSQL, AWS Cognito
Geospatial & Mapping	Leaflet, React-Leaflet, MapLibre GL, Amazon Location Service (Places & Routes)
AI & Prediction	Python 3.12, FastAPI, XGBoost, Copernicus Digital Elevation Models (DEM), Open-Meteo ERA5
External Intelligence	AWDAX Synapical Research API (State Disaster Bulletin Automated Crawler)
Authentication & Auth	Google Firebase Authentication, Capacitor Google Auth, AWS Cognito IAM Credentials
Mobile Architecture	Capacitor 8, Android SDK 34 (Debug APK with Google Services)

---

🚀 Local Development & Setup

Prerequisites

• Node.js (v18+ recommended)
• Python 3.10+ (for prediction service)
• AWS CLI configured with appropriate project credentials

Frontend Setup

# Clone the repository
git clone https://github.com/AWDAX/DOAB-Platform.git
cd DOAB-Platform/Codebase

# Install dependencies
npm install --legacy-peer-deps

# Start local development server
npm run dev
# Running on http://localhost:5173

Full Multi-Service Docker Deployment

# Run all containers (Postgres RDS proxy, FastAPI Prediction API, Frontend)
docker compose -f docker-compose.yml up --build -d

AI Prediction Service Setup

cd services/prediction-api
python3 -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate
pip install -r requirements.txt

# Build local training dataset & fit XGBoost model
python -m app.cli build-dataset
python -m app.cli train

# Launch API
uvicorn app.main:app --host 0.0.0.0 --port 8000

Building Android Native APK

# Sync web build to native Android assets
npm run build
npx cap sync android

# Build debug APK using Gradle
cd android
./gradlew assembleDebug

---

📄 Hackathon Submission Summary (AWS Architecture Breakdown)

For jury presentations and official submission forms, use this concise architectural breakdown:

### AWS Cloud Infrastructure Breakdown:
1. **Frontend Hosting & CDN**: 
   - AWS S3 (Bucket: doab-frontend-970909364340) for high-durability static asset hosting.
   - AWS CloudFront (Distribution: E20R65BKZ9U404) for global edge acceleration, SSL termination, and sub-second asset delivery.
2. **Backend & AI Inference Engine**:
   - AWS EC2 (Ubuntu 24.04 LTS) running Docker Compose containers for FastAPI prediction service, real-time elevation routing, and sensor ingestion pipelines.
3. **Database & Persistence**:
   - AWS RDS (PostgreSQL 18.3) managing relational civic disaster state, ward telemetry archives, and work order audit records.
4. **AWS High-Value Services & Emergency Mesh Used**:
   - Amazon Location Service: Pan-India forward/reverse geocoding, address autocomplete, and boundary lookup without privacy-invasive third-party trackers.
   - Amazon SNS (Simple Notification Service): Immediate SMS broadcast to Disaster Response Teams (NDRF/SDRF) and emergency contacts, plus push notifications to field crews with zero cold start.
   - Amazon SQS (Simple Queue Service): Zero-loss distress buffer queuing thousands of concurrent SOS beacons during mass flash floods.
   - AWS Lambda: Serverless emergency compute that extracts coordinates and computes nearest high-elevation relief centers on the fly.
   - AWS Cognito Identity Pools: Secure temporary credential vending for frontend Amazon Location Service SDK calls.
   - AWS CloudWatch: Container health monitoring and edge performance telemetry.

---

📚 Technical Documentation & Architecture Directory

All detailed architecture specifications, deployment guides, and design docs are consolidated under docs/:

Document	Category	Description
B2B Logistics Monetization	Commercial & Fleet	Enterprise fleet safety monetization model for Swiggy, Zepto, Zomato & 2-wheeler gig fleets
Deployment Architecture	Cloud & Infra	AWS production deployment, ECS Fargate, Amplify, and serverless disaster mesh
Geospatial Architecture	GIS & AI Engine	Elevation ridgelines, Dijkstra spatial risk engine, and hydraulic underpass models
Architecture Diagrams	Architecture	Complete system diagrams, data pipelines, and real-time telemetry flows in Mermaid
Docker & Local Setup	DevOps & Local	Full local environment bootstrap with Docker Compose & Python AI microservices
Android Native App Guide	Mobile (Android)	Capacitor native APK build, offline geolocation caching, and Android permissions
UI Design System	Design & UX	Dual-climate color tokens, neo-glassmorphism, a11y standards, and custom SVG cursors
Hackathon Video Script	Presentation	Step-by-step video demonstration flow and live product presentation script

---

👥 Engineering Team — Contrarians

• Team: Contrarians
• Project: DOAB (दोआब) — Dual-Climate Geospatial & Civic Safety Platform
• License: GNU General Public License v3.0 (GPLv3)