YOLOv8-based Smart Vision Navigation for Autonomous EV Charging Systems

Abstract

The objective of this paper is to provide advancements in charging infrastructure and automation in Electric Vehicle (EV) technology. Automating the physical connection process for electric vehicle charging remains a critical challenge due to the need for precise alignment and reliable port recognition. This study presents a comprehensive smart vision‑based navigation architecture designed for Automated Charging Robots (ACRs). The proposed approach integrates a state‑of‑the‑art YOLOv8 detector for real‑time classification of charging‑port variants with multimodal sensing combining a Three-Dimensional depth camera for spatial reconstruction and an Infrared (IR) sensor for short‑range detection to produce robust pose estimates suitable for fine manipulation. All perception and control modules are implemented within the Robot Operating System to facilitate modularity and real‑time operation. Experimental validation demonstrates that fusing depth and IR measurements reduces insertion errors in cluttered or low‑contrast scenarios. Additionally, the system includes a GSM and GPRS communication channel that transmits charging‑completion alerts to vehicle owners, thereby streamlining the user interaction loop and enhancing overall service reliability and further Smart vision-based navigation equipped with latest technologies to enable Higher efficiency in the port identification, stability, and seamless user interaction.