Description

Introduction
This project aims to design and implement a Smart Crosswalk System that leverages AI,
real-time monitoring, and renewable energy to improve pedestrian safety at non-signalized
intersections. The system will use cameras combined with AI models (YOLOv8 and pose
estimation) to detect pedestrians, especially children, approaching a crosswalk and
intending to cross. LED-embedded traffic signs powered by solar panels will serve as visual
alerts for drivers, ensuring they have sufficient time to brake.

Background
Pedestrian safety at non-signalized intersections is a major concern in urban areas. Drivers
often fail to notice pedestrians approaching crosswalks, especially at night or in poor
weather conditions. Traditional warning signs lack intelligence and real-time
responsiveness. By integrating AI-based pedestrian detection, renewable energy, and IoT
technologies, a smarter and more reliable solution can be deployed to reduce accidents and
save lives.

Project Scope
• Hardware Integration
- Install 2 AI-powered cameras in both sides of the crosswalk to monitor pedestrian activity.
- Deploy LED-based smart traffic signs around the crosswalk for visual driver alerts.
- Power the system with solar panels for energy save.
• Software Development
- Build a Python-based AI service using YOLOv8 for human detection and pose estimation to
determine if pedestrians are facing the crosswalk and within 2 meters.
- Develop a Node.js backend server to handle API routes, communication with the AI
service, and control of LED signals with LoRaWAN system.
- Implement a React.js frontend dashboard for monitoring detections, logs, and system
status.
- Store detection events and system logs in MongoDB.

• System Integration
- Establish communication between the Node.js backend and the Python AI service using
REST API.
- Send detection results from the AI service to the backend in real time.
- Trigger LED alerts through backend-controlled LoRaWAN system when pedestrians are
detected.
- Provide an admin dashboard to visualize detections and manage the system.

Student Requirements
• Familiarity with JavaScript and Python.
• Knowledge of React.js and Node.js.
• Basic understanding of MongoDB.
• Understand the concept of AI/ML frameworks (YOLOv8 and pose estimation).
• Understanding of REST API.
• Teamwork and independent learning.

Development Tools
• Frontend: React.js, Tailwind / Material UI / CSS.
• Backend: Node.js (Express.js).
• AI Service: Python (YOLOv8, OpenCV, PyTorch).
• Database: MongoDB.
• General technologies: Axios.
• Hardware Communication: LoRaWAN system for LED control.
• Additional Tools: Postman, GitHub, VSCode.

Deliverables
• AI-based Smart Crosswalk Fullstack Web App: Dashboard with real-time pedestrian
detection and system alerts.
• Specification document.
• YouTube video.
• Poster and presentation.
• Presentation.
Name: Netanel Ben Hamo
E-mail: netanelbe@hit.ac.il
AI-Based Smart Crosswalk - Fullstack Web App

AI and Machine Learning

Description

Emphasis in project execution

The project is has cooperation with the industry and combines meeting deadlines while being creative and focused on the task

THE PROJECT IS AT FULL CAPACITY
