V2X - Collision Avoidance

Abdallah Shehawey, A.O.M. Saleh, A G. K. GAMAL

# Overview The **V2X Collision Avoidance System** is an intelligent transportation platform that integrates **Vehicle-to-Everything (V2X)** communication, **embedded systems**, **artificial intelligence**, and **sensor fusion** to improve road safety through real-time collision prevention and advanced driver assistance. The platform implements a complete **V2X ecosystem**, covering all four communication domains: - 🚘 **Vehicle-to-Vehicle (V2V)** - 🚦 **Vehicle-to-Infrastructure (V2I)** - 🚶 **Vehicle-to-Pedestrian (V2P)** - ☁️ **Vehicle-to-Network (V2N)** --- # System Architecture ## 🚘 Vehicle-to-Vehicle (V2V) The V2V subsystem is implemented on the **STM32F446RE** using **Bare-Metal C** and **FreeRTOS**. It executes multiple cooperative ADAS algorithms by combining information received from neighboring vehicles with onboard sensor data to detect hazards and generate real-time safety alerts. ### Implemented ADAS Features - Forward Collision Warning (**FCW**) - Electronic Emergency Brake Lights (**EEBL**) - Blind Spot Warning (**BSW**) - Do Not Pass Warning (**DNPW**) - Intersection Movement Assist (**IMA**) - Safe Distance Warning (**SDW**) --- ## 🚦 Vehicle-to-Infrastructure (V2I) The V2I subsystem runs on a **Raspberry Pi**, enabling communication with intelligent traffic infrastructure. Computer vision is powered by **YOLOv8** to perform: - Traffic signal detection - Vehicle distance estimation - License plate recognition This enables vehicles to understand road conditions and traffic signal status in real time. --- ## 🚶 Vehicle-to-Pedestrian (V2P) The V2P subsystem utilizes an **ONNX-based AI pedestrian detection model** running on the Raspberry Pi. Unlike traditional V2P solutions, pedestrian detection is performed entirely through **computer vision**, eliminating the need for pedestrians to carry smartphones or dedicated communication devices. --- ## ☁️ Vehicle-to-Network (V2N) Cloud connectivity is provided through **MQTT**, enabling: - Real-time telemetry transmission - Hazard notifications - Remote monitoring - Future Smart City integration --- # Communication Layer Wireless communication between vehicles and infrastructure is handled by **ESP32** modules using: - ESP-NOW - Wi-Fi An **IPC Hub** running on the Raspberry Pi acts as the central message broker, routing data between: - V2V - V2I - V2P - V2N - ADAS Dashboard - Control Server --- # ADAS Dashboard A modern web-based **ADAS Dashboard** was developed using: - HTML - CSS - JavaScript with a **Python WebSocket Server** for real-time communication. ### Dashboard Features - Live vehicle telemetry - Radar visualization - Object detection results - Real-time safety alerts - Vehicle status monitoring Safety alerts are classified into multiple levels: - 🟢 Safe - 🟡 Warning - 🔴 Critical --- # Control Server The platform includes a dedicated **Control Server** capable of issuing emergency commands. When a critical collision risk is detected, the server can automatically trigger emergency actions such as: - Vehicle stop command - Emergency braking request - Hazard notification broadcasting --- # Sensor Fusion To improve environmental perception and decision accuracy, the platform combines multiple sensing technologies: - Ultrasonic sensors - IMU measurements - V2X communication - AI-based computer vision This unified sensor fusion architecture enables accurate, low-latency collision avoidance decisions in real time. --- # Technologies Used ## Embedded Systems - STM32F446RE - Bare-Metal C - FreeRTOS - ESP32 - Raspberry Pi ## Artificial Intelligence - YOLOv8 - ONNX Runtime - Computer Vision ## Communication - ESP-NOW - Wi-Fi - MQTT - WebSockets ## Web Technologies - HTML - CSS - JavaScript - Python --- # Project Highlights - Complete implementation of all four V2X communication domains - Multi-node embedded architecture - Real-time cooperative ADAS algorithms - AI-powered object and pedestrian detection - Lo