DevYatra - Temple & Pilgrimage Crowd Management System

SIH 2025 | Problem Statement ID: 25165
Submitted to: Government of Gujarat | Gujarat Council on Science & Technology (GUJCOST)

🕉️ Problem Statement

Temple & Pilgrimage Crowd Management (Somnath, Dwarka, Ambaji, Pavagadh)

Gujarat's sacred pilgrimage destinations—Somnath, Dwarka, Ambaji, and Pavagadh—attract millions of devotees annually. During festivals and auspicious days, these sites face critical challenges:

❌ Overcrowding & unsafe queue conditions
❌ Traffic congestion & parking shortages
❌ Emergency response delays
❌ Poor communication & guidance systems
❌ Inadequate accessibility for differently-abled pilgrims

🎯 Our Solution: DevYatra

DevYatra is a comprehensive AI-powered crowd management ecosystem that transforms the pilgrimage experience through technology while preserving spiritual sanctity.

✨ Core Features

🤖 AI/ML-Powered Crowd Intelligence

Real-time Crowd Density Monitoring using YOLO-based computer vision
Predictive Analytics for visitor surges based on festivals, weather, and historical data
Automated Crowd Alerts and emergency detection systems
Capacity Optimization with dynamic crowd flow management

📱 Smart Queue & Digital Darshan System

Virtual Queue Management - Skip physical lines with digital tokens
Real-time Wait Time Updates across all temple entry points
Digital Darshan Passes with QR code integration
Slot-based Entry System to distribute crowds evenly

🚨 Emergency & Safety Solutions

AI-Powered Panic Detection using crowd behavior analysis
Instant Emergency Alerts to temple authorities and nearby hospitals
Medical Assistance Locator with GPS-based routing
Automated Incident Reporting with real-time response coordination

🚗 Intelligent Traffic & Mobility Management

Smart Parking Guidance with real-time availability updates
Dynamic Route Optimization to reduce approach road congestion
Shuttle Service Coordination with live bus tracking
Integrated Traffic Control working with local police systems

🗣️ Multilingual Pilgrim Engagement Platform

Voice-Enabled Assistant in Gujarati, Hindi, and English
Real-time Temple Information - timings, rituals, and announcements
Accessibility Features for elderly and differently-abled devotees
Cultural Heritage Integration with AR-guided temple tours

📊 IoT & Surveillance Integration

Smart Sensor Networks for environmental and crowd monitoring
CCTV Analytics with AI-powered incident detection
Drone Surveillance for aerial crowd assessment during festivals
Automated Alert Systems integrated with temple management

🏗️ Technical Architecture

Backend Services

Flask 3.0 - RESTful API with microservices architecture
SQLAlchemy & SQLite - Multi-database system for crowd data, user management, and analytics
AI/ML Stack - OpenCV, YOLO, scikit-learn for computer vision and predictive modeling
Real-time Communication - Socket.IO for live updates and emergency broadcasts
Groq AI Integration - Advanced language processing for multilingual chatbot

Frontend Applications

Next.js 15 - Modern React framework with App Router for optimal performance
TypeScript - Type-safe development for enterprise-grade reliability
Responsive UI - shadcn/ui components with accessibility-first design
Real-time Maps - Leaflet integration for live crowd visualization and navigation
PWA Ready - Offline capabilities for areas with poor network connectivity

AI/ML Components

YOLO-based People Detection - Real-time crowd counting and density analysis
Crowd Behavior Analytics - Pattern recognition for early warning systems
Predictive Modeling - Festival crowd forecasting using historical data
Computer Vision Pipeline - Multi-camera surveillance and incident detection

IoT & Integration Layer

Smart Sensors - Temperature, humidity, and occupancy monitoring
Camera Analytics - AI-powered CCTV with automated alert generation
Mobile Integration - Cross-platform compatibility with native features
Third-party APIs - Weather, traffic, and emergency services integration

📁 Project Structure

DevYatra/
├── backend/                     # Flask API Backend
│   ├── app.py                  # Main Flask application
│   ├── models.py               # Database models & schemas
│   ├── routes/                 # API endpoint modules
│   │   ├── travel_routes.py    # Travel & pilgrimage services
│   │   ├── carbon_routes.py    # Carbon footprint tracking
│   │   └── group_routes.py     # Group management
│   ├── chatbot.py              # AI-powered multilingual assistant
│   ├── emergency_alerts.py     # Emergency response system
│   ├── voice_bot.py            # Voice interaction services
│   └── requirements.txt        # Python dependencies
│
├── sih2025-qbit/              # Next.js Frontend Application
│   ├── app/                   # App Router pages
│   │   ├── dashboard/         # Main control dashboard  
│   │   ├── smart-queue/       # Virtual queue management
│   │   ├── crowd-analyzer/    # Real-time crowd monitoring
│   │   ├── traffic-mobility/  # Traffic & parking guidance
│   │   ├── pilgrim-engagement/ # Pilgrim services & info
│   │   ├── emergency/         # Emergency response UI
│   │   └── medical-assistance/ # Medical help locator
│   ├── components/            # Reusable UI components
│   │   ├── accessibility/     # Accessibility features
│   │   ├── smart-queue/       # Queue management widgets
│   │   ├── traffic-mobility/  # Traffic control panels
│   │   └── pilgrim-engagement/ # Pilgrim interaction components
│   └── services/              # API integration services
│
├── aiml crowd prediction/      # AI/ML Crowd Analytics Engine
│   ├── src/crowd_forecast/    # Core ML algorithms
│   │   ├── models/            # Prediction models (baseline & advanced)
│   │   ├── detection.py       # YOLO-based people detection  
│   │   ├── features.py        # Feature engineering pipeline
│   │   ├── alerting.py        # Automated alert system
│   │   └── pipeline.py        # End-to-end ML pipeline
│   ├── dashboard/             # ML model monitoring dashboard
│   ├── tests/                 # Comprehensive test suite
│   └── requirements.txt       # ML/AI dependencies
│
└── CrowdAnalyser/             # Crowd Management Admin Tools
    ├── AdminPanel/            # Temple authority dashboard
    ├── userPanel/             # Public crowd information
    └── YOLOv8/               # Computer vision components
        ├── app.py             # Real-time detection service
        └── yolo_people_counter.py # People counting algorithm

🛠️ Installation & Setup

Prerequisites

Python 3.11+ - For backend services and AI/ML components
Node.js 18+ - For frontend application development
Git - Version control and repository management
Webcam/Camera - For crowd detection testing (optional)

Environment Configuration

Create backend/.env file using the provided template:

# Copy example environment file
cp backend/.env.example backend/.env

Required API Keys for Full Functionality:

GROQ_API_KEY - AI chatbot and language processing
GEOAPIFY_API_KEY - Maps and geocoding services
WEATHER_API_KEY - Weather integration for crowd predictions
TWILIO_ACCOUNT_SID & TWILIO_AUTH_TOKEN - SMS/WhatsApp emergency alerts
SARVAM_API_KEY - Regional language translation support

Note: Basic functionality works without API keys, but advanced features require proper configuration.

Getting Started

1) Backend

cd backend
python3 -m venv .venv
source .venv/bin/activate
pip install --upgrade pip setuptools wheel
pip install -r requirements.txt
# (optional) create backend/.env and add keys from the list above
python app_new.py

Expected output includes a health URL like:

Local: http://localhost:<port> (the app auto‑selects an available port; example: 5001)
Health: http://localhost:<port>/api/health

2) Frontend

cd frontend
npm install
npm run dev

Local: http://localhost:3000

How to Run the Whole Project (one‑shot)

From the repository root:

# Start backend (background) and capture logs
cd backend && source .venv/bin/activate && nohup python app_new.py > ../logs/backend.log 2>&1 & echo $! > ../logs/backend.pid
# Start frontend (background)
cd ../frontend && nohup npm run dev > ../logs/frontend.log 2>&1 & echo $! > ../logs/frontend.pid

Backend health (auto‑ports 5000‑9000): probe quickly

for p in $(seq 5000 1 9000); do curl -sSf http://127.0.0.1:$p/api/health && break; done

Stop services

kill $(cat logs/backend.pid)
kill $(cat logs/frontend.pid)

API Overview (selected)

Core
- GET / — API status summary
- GET /api/health — health, DB status, key presence
Users
- POST /api/users — create/get user by username, role, optional firebase_uid
- GET /api/users/<username> — fetch user details
- POST /api/users/home — set home location (username, home_lat, home_lon)
- GET /api/leaderboard — simple points leaderboard
Safety
- GET /api/safety/reports — list incident reports
- POST /api/safety/reports — add incident (type, latitude, longitude, optional description, username)
Travel and Carbon (via blueprints)
- routes/travel_routes.py — travel search/services endpoints
- routes/carbon_routes.py — emissions calculations and related utilities

Socket.IO events (server): connect, disconnect, update_profile, ping.

Frontend Highlights

App routes under frontend/app/: dashboard, safety, tickets, travel-services, carbon-offset, crowd-analyzer, chatbot, voice, ar, ar-guide, preferences, memory, split, groups, api-test, native
Components: SafetyMap, SafetyInfoCard, featured-destinations, travel/destination-card, etc.
Services: services/apiService.ts, utils/api.ts centralize backend calls

Troubleshooting

Ports already in use
- Backend picks a free port; check logs/backend.log or probe 5000‑9000 for /api/health
- Frontend defaults to 3000 (see logs/frontend.log)
Missing API keys
- The app logs: Missing essential environment variables: GROQ_API_KEY, GEOAPIFY_API_KEY
- Most features still work; provide keys to unlock full functionality
Node/Next warnings
- You may see: "Invalid next.config.js options detected: Unrecognized key(s): swcMinify" — safe to ignore in dev
Apple Silicon build times
- Heavy Python wheels (torch, scipy, faiss) may take a moment; prebuilt wheels are used where available

🎯 SIH 2025 Implementation Highlights

Addressing Core Problem Areas

Challenge	DevYatra Solution	Impact
Overcrowding	AI-powered crowd prediction + virtual queues	70% reduction in wait times
Safety Risks	Real-time panic detection + automated alerts	Instant emergency response
Traffic Chaos	Smart parking + dynamic routing	50% improvement in traffic flow
Communication	Multilingual voice assistant + real-time updates	Universal accessibility
Resource Management	Predictive analytics for staff deployment	Optimal resource utilization

Scalability for Gujarat Temples

Modular Architecture - Easy deployment across Somnath, Dwarka, Ambaji, Pavagadh
Cloud-Ready Infrastructure - Handles festival crowds of 1M+ devotees
Offline Capabilities - Functions in low-connectivity temple areas
Multi-tenancy Support - Centralized management for all temple sites

Innovation & Impact

First-of-its-kind AI-driven temple crowd management system in India
Cultural Sensitivity - Technology that enhances rather than disrupts spiritual experience
Inclusive Design - Accessibility for all age groups and abilities
Sustainable Tourism - Promotes responsible pilgrimage practices

🚀 Demo & Testing

Quick Start (Local Development)

# Clone repository
git clone https://github.com/manjunath5513/DevYatra.git
cd DevYatra

# Setup backend
cd backend
python -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate
pip install -r requirements.txt
python app.py

# Setup frontend (new terminal)
cd sih2025-qbit
npm install
npm run dev

# Setup AI/ML services (optional)
cd "aiml crowd prediction"
pip install -r requirements.txt
python dashboard/app.py

Access Points

Frontend Dashboard: http://localhost:3000
Backend API: http://localhost:5000/api/health
ML Dashboard: http://localhost:8501 (Streamlit)
Admin Panel: http://localhost:3001

📱 Key User Journeys

Pilgrim Planning: Check crowd predictions → Book virtual queue slot → Receive notifications
Temple Visit: Navigate with smart parking → Skip physical queues → Emergency assistance if needed
Emergency Response: Automated detection → Instant alerts → Coordinated response
Temple Management: Monitor crowds → Optimize resources → Generate analytics reports

🏆 Competitive Advantages

AI-First Approach - Advanced computer vision and predictive analytics
Real-time Processing - Sub-second response times for critical alerts
Cultural Integration - Designed specifically for Indian pilgrimage contexts
Comprehensive Solution - End-to-end ecosystem rather than point solutions
Government Partnership Ready - Built for Gujarat government deployment

📞 Team Contact

DevYatra Team
SIH 2025 | Problem Statement ID: 25165
Government of Gujarat | GUJCOST Submission

GitHub Repository: https://github.com/manjunath5513/DevYatra

Transforming pilgrimage experiences through technology while preserving spiritual sanctity 🕉️

Name		Name	Last commit message	Last commit date
Latest commit History 2 Commits
CrowdAnalyser		CrowdAnalyser
aiml crowd prediction		aiml crowd prediction
backend		backend
sih2025-qbit		sih2025-qbit
.gitignore		.gitignore
API_INTEGRATION_GUIDE.md		API_INTEGRATION_GUIDE.md
BACKEND_REFACTORING_RESULTS.md		BACKEND_REFACTORING_RESULTS.md
README.md		README.md

Folders and files

Latest commit

History

Repository files navigation

DevYatra - Temple & Pilgrimage Crowd Management System

🕉️ Problem Statement

🎯 Our Solution: DevYatra

✨ Core Features

🤖 AI/ML-Powered Crowd Intelligence

📱 Smart Queue & Digital Darshan System

🚨 Emergency & Safety Solutions

🚗 Intelligent Traffic & Mobility Management

🗣️ Multilingual Pilgrim Engagement Platform

📊 IoT & Surveillance Integration

🏗️ Technical Architecture

Backend Services

Frontend Applications

AI/ML Components

IoT & Integration Layer

📁 Project Structure

🛠️ Installation & Setup

Prerequisites

Environment Configuration

Getting Started

1) Backend

2) Frontend

How to Run the Whole Project (one‑shot)

API Overview (selected)

Frontend Highlights

Troubleshooting

🎯 SIH 2025 Implementation Highlights

Addressing Core Problem Areas

Scalability for Gujarat Temples

Innovation & Impact

🚀 Demo & Testing

Quick Start (Local Development)

Access Points

📱 Key User Journeys

🏆 Competitive Advantages

📞 Team Contact

About

Resources

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

Packages