EZGO – Smart Mobility Infrastructure for Tunisia

EZGO is a dual-platform mobile and web solution designed to modernize Tunisia's public transportation system. It delivers real-time navigation, smart reservations, and data-driven operational tools that improve reliability for riders and profitability for transport operators.

Features

For Passengers & Tourists

• Seamless mobile app with live bus tracking
• Accurate ETAs and route visualization
• Seat availability and reservations
• Digital tickets and booking history
• Driver information and ratings

For Bus Companies

• Complete fleet management system
• Driver assignment and scheduling
• Real-time operations monitoring
• Profitability dashboard and analytics
• Performance insights and optimization

For Administrators

• Centralized user and company management
• System-wide analytics and reporting
• Permissions and roles management
• Quality assurance and compliance tools

Architecture

This project is a Turborepo monorepo containing three applications:

ezgo/
├── apps/
│   ├── web/          # Next.js Admin & Company Dashboard
│   ├── api/          # NestJS Backend API
│   └── mobile/       # Expo React Native Mobile App
├── packages/         # Shared packages (future)
└── docker-compose.yml

Tech Stack

Web Dashboard (apps/web)

• Framework: Next.js 16 (App Router)
• Styling: TailwindCSS v4
• UI Components: Shadcn UI
• State Management: Tanstack React Query
• Language: TypeScript

Backend API (apps/api)

• Framework: NestJS
• Database: PostgreSQL (via Docker)
• ORM: Prisma 7
• Validation: Zod
• Language: TypeScript

Mobile App (apps/mobile)

• Framework: Expo (React Native)
• Styling: NativeWind (TailwindCSS for React Native)
• Navigation: React Navigation
• Language: TypeScript

Tooling

• Monorepo: Turborepo
• Linting/Formatting: Biome.js
• Package Manager: npm

Installation

Prerequisites

• Node.js 18+ and npm
• Docker and Docker Compose
• Git

Setup

1. Clone the repository

   git clone <repository-url>
   cd EzGo

2. Install dependencies

   npm install

3. Start PostgreSQL database

   docker-compose up -d

4. Setup environment variables

   Create apps/api/.env:

   DATABASE_URL="postgresql://postgres:password@localhost:5432/ezgo?schema=public"

5. Generate Prisma Client

   cd apps/api
   npx prisma generate
   npx prisma migrate dev --name init
   cd ../..

Running the Applications

Development Mode

Run all applications concurrently:

npm run dev

Or run individually:

Web Dashboard:

cd apps/web
npm run dev
# Access at http://localhost:3000

API Server:

cd apps/api
npm run start:dev
# Access at http://localhost:3001

Mobile App:

cd apps/mobile
npm start
# Scan QR code with Expo Go app

Production Build

npm run build

Screenshots

Web Dashboard

Landing Page

The landing page provides a clean introduction to EZGO with quick access to the dashboard.

Admin Dashboard

The admin dashboard displays key metrics including:

• Total Revenue with growth indicators
• Active Buses count
• Active Users statistics
• Real-time activity monitoring
• Overview charts and recent sales

Mobile App

The mobile app features:

• Passenger Mode: Real-time bus tracking, seat booking, and trip history
• Driver Mode: Assigned routes, passenger counts, and navigation assistance

Database Schema

The application uses the following main entities:

• User: Passengers, Drivers, Company Admins, and System Admins
• Company: Bus companies managing fleets
• Bus: Fleet vehicles with capacity and status
• Route: Bus routes with stops and coordinates
• Trip: Scheduled trips with real-time status
• Booking: Passenger reservations and tickets

AI Features

Fatigue Detection

The fatigue-detection.ipynb notebook implements driver fatigue detection using computer vision. This notebook must be run in the Kaggle environment.

Dataset: Drowsiness Dataset

Setup Instructions:

1. Upload the notebook to Kaggle
2. Add the dataset using the URL above
3. Run the notebook in Kaggle's environment

Suspicious Behaviour Detection

The suspicious_behaviour_detection.ipynb notebook implements suspicious activity detection for passenger safety monitoring. This notebook is trained on a custom dataset and should be run on Kaggle.

Dataset: Suspicious Activity Detection

Setup Instructions:

1. Upload the notebook to Kaggle
2. Set up the dataset from Roboflow
3. Run the notebook in Kaggle's environment

Development

Code Quality

The project uses Biome.js for linting and formatting:

# Format code
npm run format

# Check code quality
npm run check

VSCode Setup

Recommended extensions are configured in .vscode/extensions.json:

• Biome
• Prisma
• Tailwind CSS IntelliSense
• Expo Tools

API Endpoints

Companies

• GET /company - List all companies
• GET /company/:id - Get company by ID
• POST /company - Create new company
• PATCH /company/:id - Update company
• DELETE /company/:id - Delete company

Additional endpoints for Users, Buses, Routes, Trips, and Bookings will be added.

Roadmap

• Complete API endpoints for all entities
• Implement authentication and authorization
• Add real-time tracking with WebSockets
• Integrate payment gateway
• Add push notifications
• Implement analytics dashboard
• Add map integration (Google Maps/Mapbox)
• Create mobile app screens for all features
• Add unit and integration tests
• Integrate IoT with AI
• Deploy to production