Vision-Based Robotic System for Smart EV Charging Using Deep Learning

Electric Vehicles (EVs) are increasingly favored worldwide due to their environmental benefits and advancements in battery technology, leading to lower costs. However, challenges such as limited range and inadequate charging infrastructure hinder their widespread adoption. Establishing efficient charging stations is crucial to support transport electrification and meet the rising demand for EV charging. This abstract provides an overview of EV charging, covering different charging settings, standards, and emerging technologies. Key challenges include standardization, interoperability, and the charging imbalance effect on the electricity grid, which can lead to fluctuations in power demand. Addressing these issues requires innovative solutions that enhance charging efficiency and grid stability. To overcome these challenges, an automated robotic charging system is proposed. This work presents a 3D vision-based automated charging solution using YOLOv7 for accurate detection and alignment of EV charging ports, addressing connector and location diversity in real-world scenarios. This foundational work is a stepping stone toward fully autonomous and scalable robotic EV charging systems.