Campus Navigator

An intelligent campus navigation system that helps students, visitors, and staff find buildings, departments, and popular spots on campus quickly and easily.

---

Table of Contents

• Problem Statement
• Solution
• Features
• Tech Stack
• Architecture
• Installation
• API Endpoints
• Project Structure

---

Problem Statement

Challenges Faced by Campus Visitors & Students

Challenge	Description
Navigation Difficulty	New students, visitors, and parents struggle to find specific buildings, departments, or facilities on large college campuses
Lack of Real-time Assistance	Traditional campus maps are static and don't provide interactive guidance
Information Accessibility	Finding information about department locations, seminar halls, hostels is fragmented
No Personalized Help	Users cannot ask natural language questions and get immediate responses
Mobile Accessibility	Existing solutions often don't work well on mobile devices

---

Solution

Campus Navigator is a full-stack web application that provides:

Solution	Benefit
Interactive Map	Visual navigation with Leaflet/Google Maps
Smart Search	Autocomplete-enabled search for locations
AI Chatbot	Natural language queries using semantic similarity
Popular Places	Dynamic tracking of frequently searched locations
AR Navigation	Augmented reality overlays for real-world navigation
Mobile-Responsive	Works on phones, tablets, and desktops

---

Features

Interactive Campus Map

• Leaflet maps with campus overlay
• Real-time GPS location tracking
• Visual markers for all campus buildings
• Turn-by-turn navigation

Smart Search

• Autocomplete suggestions as you type
• Fuzzy matching for location names
• Debounced API calls for performance

AI Chatbot Assistant

• Semantic similarity matching using Sentence Transformers
• all-MiniLM-L6-v2 model (384-dim embeddings)
• AIML pattern matching fallback
• Context-aware responses

AR Navigation

• Camera-based augmented reality view
• Compass heading for direction
• Distance overlay to destination
• Real-time position updates

Popular Places

• Tracks most searched locations
• Dynamic ranking based on user interest
• Quick access buttons

---

Tech Stack

Frontend

Technology	Purpose
React.js 18	UI Framework
React Router v6	Navigation
Axios	HTTP Client
Leaflet	Map Rendering
CSS3	Styling

Backend

Technology	Purpose
Flask 3.x	Web Framework
Flask-CORS	Cross-Origin Requests
Gunicorn	Production Server
Transformers	NLP Models
PyTorch	ML Framework
AIML	Pattern Chatbot

AI/ML Components

Component	Technology
Semantic Search	sentence-transformers/all-MiniLM-L6-v2
Embeddings	384-dimensional vectors
Similarity	Cosine Similarity (threshold: 0.7)
Lazy Loading	Memory optimization

---

Architecture

                    FRONTEND (React.js)                       
         
   Navbar     Search     MapPage    Chatbot/AR Nav   
         
                  
                     Axios HTTP Client                        

                              HTTP/HTTPS

                    BACKEND (Flask)                           
   
                REST API Endpoints                          
    /api/search    /api/top-places    /chat              
   
                                                           
      
   JSON Database   Location       Chatbot Module       
   (locations,     Service          
    searches)                   Semantic Model       
      (Transformers)       
                                        
                                    AIML Patterns        
                                        
                                    

---

Installation

Prerequisites

• Node.js 16+ and npm
• Python 3.9+
• Git

Backend Setup

cd backend
python -m venv venv

# Windows
.\venv\Scripts\activate

# Linux/Mac
source venv/bin/activate

pip install -r requirements.txt
python main.py

Backend runs at: http://localhost:5001

Frontend Setup

cd frontend
npm install
npm start

Frontend runs at: http://localhost:3000

---

API Endpoints

Method	Endpoint	Description
GET	/api/search?q={query}	Search locations
POST	/api/search	Record search (update popularity)
GET	/api/top-places	Get top 5 popular places
POST	/chat	Send message to AI chatbot

Example Requests

# Search locations
curl "http://localhost:5001/api/search?q=computer"

# Get popular places
curl "http://localhost:5001/api/top-places"

# Chat with bot
curl -X POST "http://localhost:5001/chat" \
  -H "Content-Type: application/json" \
  -d '{"message": "Where is the library?"}'

---

Project Structure

campus-navigator/
 backend/
    main.py                 # Flask entry point
    config.py               # Configuration
    requirements.txt        # Python dependencies
    Dockerfile              # Container config
    Procfile                # Deployment config
    chatbot/
       chat.py             # AIML handler
       bot.aiml            # Chat patterns
       semantic_model.py   # ML semantic search
       questions.json      # FAQ knowledge base
    *.json                  # Location databases

 frontend/
    src/
       App.js              # Main component
       Components/
          Navbar.js       # Navigation + search
          Search.js       # Autocomplete search
          MapPage.js      # Map + navigation
          chatbot.js      # AI chat interface
          ARScene.js      # AR navigation
          ARNavigation.js # AR overlay
       api/
           geolocation.js  # GPS utilities
    package.json

 README.md

---

Implementation Highlights

Semantic Search Algorithm

# 1. Load pre-trained model (lazy loaded)
model = "sentence-transformers/all-MiniLM-L6-v2"

# 2. Generate 384-dim embedding for query
user_embedding = get_embedding(user_query)

# 3. Compare with FAQ using cosine similarity
for question in FAQ:
    similarity = cosine_similarity(user_embedding, question_embedding)
    if similarity > 0.7:
        return get_answer(question)

# 4. Fallback to AIML patterns
return aiml_response(user_query)

Memory Optimization

• Lazy loading: ML models load only when chatbot is used
• Single worker: Gunicorn uses 1 worker for 512MB limit
• Startup: ~100MB After chatbot use: ~300MB

---

Future Enhancements

Feature	Description
Voice Commands	Voice-based search
Indoor Mapping	Floor-by-floor navigation
Event Integration	Campus events & schedules
Multi-language	Regional language support
Accessibility	Screen reader support

---