Fuel Efficient Self-Driven Vehicle using CNN with V2V Communication



In the age of Artificial Intelligence (AI), transportation faces a revolution using AI and computer vision via cameras, ultrasonic sensors for obstacle detection, GPS for location, and ADXL345 as a compass. The goal is to enhance vehicle safety and fuel efficiency using Machine Learning (ML) with the Carla Driving Simulator’s Lane Detection dataset and Vehicle-to-Vehicle (V2V) communication. Raspberry Pi and Arduino handle calculations, enabling AI predictions with models like Convolutional Neural Network (CNN), You Only Look Once (YOLO), and OpenCV. Grayscale conversion and Canny edge detection reduce channels, while YOLO and CNN perform image segmentation. ML guides the vehicle to find the shortest, safest route autonomously, integrating navigation, obstacle avoidance, anti-collision systems, GPS, and fuel efficiency. Adherence to government road rules enhances passenger safety and enables communication with owners for service alerts and issue identification. Location-specific updates optimize vehicle performance.


Keywords: Computer Vision, CNN, Obstacle Avoidance, Anti-Collision, GPS, Fuel-Efficient


Conference Name: 10th international Conference on Business Analytics and Intelligence (2023- ICBAI)


Vignesh Krishnamurthy

Soujanya Kelaginamani

Paramesh G

Dr. Shinu Abhi

Leave a Reply

Your email address will not be published. Required fields are marked *