Pravah (प्रवाह)

Data Driven Real-Time Delhi Flood Risk Monitoring & Management System

Live Demo: https://pravah-proj.vercel.app/

A comprehensive, data-driven dashboard designed to help city administrators monitor, predict, and manage urban flood risks in real-time. It integrates rainfall data, drainage capacity, and citizen complaints to generate dynamic risk scores for 250 municipal wards in Delhi.

✨ Features

• Real-Time Risk Engine — Dynamic algorithm calculating risk scores (0-100) based on rainfall (mm), drainage capacity (%), and active complaints.
• Interactive City Map — Visualizes 250 wards with color-coded risk markers (Red/Yellow/Green) and pulse animations for high-risk zones.
• Dual-Role Interface —

1. Citizen Portal: Submit geo-tagged waterlogging complaints bala
2. Admin Dashboard: Monitor KPIs, view ward-level analytics, and update infrastructure status.

• Professional Visualization — Clean, government-grade UI with data-dense tables and actionable insights.
• Dual Themes — Both light mode and dark mode implemented to suit all users.
• Simulated IoT Data — Realistic, synchronized datasets for Rainfall and Drainage across all Delhi zones.

🏗️ Architecture

┌──────────────┐      ┌──────────────┐      ┌──────────────┐
│  IoT Sensors │      │   Citizen    │      │    Admin     │
│ (Rain/Drain) │      │  Reporting   │      │   Actions    │
└──────┬───────┘      └──────┬───────┘      └──────┬───────┘
       │                     │                     │
       ▼                     ▼                     ▼
┌──────────────────────────────────────────────────────────┐
│                   Data Aggregation Layer                 │
│              (CSV / In-Memory State Store)               │
└────────────────────────────┬─────────────────────────────┘
                             │
                             ▼
┌──────────────────────────────────────────────────────────┐
│                    Risk Engine (Python)                  │
│   Formula: (Drainage * 0.5) + (Rain * 0.3) + (Logs * 0.2)│
└────────────────────────────┬─────────────────────────────┘
                             │
                             ▼
┌──────────────┐      ┌──────────────┐      ┌──────────────┐
│   REST API   │◀───▶│   Frontend   │◀ ──▶│ Leaflet Maps │
│  (FastAPI)   │      │  (HTML/JS)   │      │ (Visuals)    │
└──────────────┘      └──────────────┘      └──────────────┘

📁 Project Structure

Pravah/
├── backend/
│   ├── data/
│   │   ├── wards.csv           # 250 Ward Geo-coordinates & Zones
│   │   ├── rainfall.csv        # Live Rainfall Data (mm)
│   │   ├── drainage.csv        # Infrastructure Capacity (%)
│   │   └── complaints.csv      # Citizen Report Log
│   │
│   ├── src/
│   │   ├── main.py             # FastAPI Entry Point & CORS
│   │   ├── config.py           # Settings & Constants
│   │   ├── models.py           # Pydantic Schemas
│   │   └── services/
│   │       ├── api.py          # API Route Handlers
│   │       ├── data_loader.py  # CSV Parsing & Merging Logic
│   │       ├── risk_engine.py  # Risk Calculation Algorithm
│   │       └── admin_ops.py    # Admin Action Logic
│   │
│   └── requirements.txt        # Python Dependencies
│
├── frontend/
│   ├── css/
│   │   ├── styles.css          # Main Theme & Responsive Rules
│   │   ├── admin.css           # Dashboard Specific Styles
│   │
│   ├── js/
│   │   ├── common.js           # API Config & Utilities
│   │   ├── admin.js            # Dashboard Logic & Maps
│   │   ├── dashboard.js        # Citizen Map & Reporting
│   │   ├── complaints.js       # Form Handling
│   │   └── login.js            # Auth Logic
│   │
│   ├── index.html              # Landing Page
│   ├── admin.html              # Main Admin Control Room
│   ├── dashboard.html          # Citizen Reporting Page
│   └── login.html              # Admin Login Gate
│   └── about.html              # About The Project
│
└── README.md

🚀 Quick Start

Prerequisites

• Python 3.9+
• Web Browser(Desktop Mode for now)
• Git

1. Clone & Setup

# Clone the repository
git clone https://github.com/prabhatbhatiaa/pravah.git
cd pravah

# Backend Setup
cd backend
python -m venv venv
.\venv\Scripts\activate  # Windows
# source venv/bin/activate # Mac/Linux

# Install dependencies
pip install -r requirements.txt

2. Runs the Backend

# Start Uvicorn Server
uvicorn src.main:app --reload

3. Run the Frontend

1. Open the frontend folder.
2. Open index.html via Live Server (VS Code Extension) or simply double-click it.
3. Important: Ensure frontend/js/common.js points to http://localhost:8000 for local development.

🔐 System Access

1. Landing Page — : Public access to project overview.
2. About Us Page — : Public accessto the details of the project.
3. Citizen Dashboard — Public access to view map & report issues.
4. Admin Portal — : Restricted access via Session Storage.

📋 API Endpoints

Public

• GET /api/wards — Returns full GeoJSON of all 250 wards with calculated risk.
• GET /api/risk-summary — Returns aggregate stats (High Risk count, Total Wards).
• POST /api/complaint — Submit a new citizen complaint.

Admin

• GET /api/admin/overview — Returns detailed tabular data sorted by risk.
• POST /api/admin/update-drainage — Update drainage capacity for a specific ward.

🧠 Risk Calculation Logic

The Risk Engine (risk_engine.py) calculates a score (0-100) for every ward in real-time

1. Drainage Factor (50%): Inverse of capacity. Lower drainage = Higher risk.
2. Rainfall Factor (30%): Normalized against a critical threshold (150mm).
3. Complaint Factor (20%): Active complaints boost the risk score significantly.

Final Score = (Drainage_Risk * 0.5) + (Rainfall_Risk * 0.3) + (Complaint_Risk * 0.2)

🛠️ Technology Stack

Backend:

• FastAPI — High-performance Python framework.
• Pandas — Data manipulation and CSV handling.
• Uvicorn — ASGI Server.

Frontend:

• HTML5 / CSS3 — Responsive layout with CSS Variables.
• Vanilla JavaScript — Logic without heavy frameworks.
• Leaflet.js — Interactive Maps & Marker rendering.

Deployment:

• Render — Python Backend Hosting.
• Vercel — Static Frontend Hosting.

📦 Dependencies

fastapi
uvicorn
pandas
pydantic
python-multipart

Create backend/requirements.txt with the above and run:

pip install -r backend/requirements.txt

🔒 Security Considerations

• Session Storage Auth — Basic client-side session management for the hackathon prototype.
• CORS Configuration — Backend restricted to specific origins in production.

🐛 Troubleshooting

Map not loading:

• Check if the Backend URL in frontend/js/common.js matches your running server.
• Ensure no trailing slash in the API base URL (e.g., https://api.com not https://api.com/).

Admin page redirects to login:

• Ensure you actually logged in via login.html to set the session key.
• Check browser console for sessionStorage errors.

High latency on first load:

• The backend is hosted on a free Render instance which spins down after inactivity. The first request might take 50 seconds to wake it up.

🎯 Roadmap

• Integration with live IMD Weather API.
• IoT Sensor integration for real-time water level monitoring.
• SMS Alerts to citizens in High-Risk zones.
• Predictive AI model for 24hr flood forecasting.
• Mobile-responsive UI improvements

👤 Author

Prabhat Bhatia & Suhani Yadav

---

Note: This project was built for a govt level hackathon. The data for "Rainfall" and "Drainage" is simulated based on real-world Delhi ward boundaries and flood hotspots.