Autonomous control of shore robotic charging systems based on computer vision

Abstract

Recently, there has been a rapid increase in electric vessels as in electric land vehicles. Since vessels such as ferries and ships operate on a specific schedule, the time to be spent for charging on the shore is limited. The charging process involves connecting equipment, transferring energy, and disconnecting equipment. The faster and easier connection and disconnection process is, the more time is for energy transfer. This paper introduces a novel autonomous robotic charging system that combines Programmable Logic Controller (PLC), embedded platform, camera, and lasers to detect the vessel's charging socket in the charging area by computer vision. Initially, the embedded platform detects the charging socket and position of the electric vessel with the camera using You-Only-Look-Once (YOLOv5). After that, when the charging socket is detected, laser sensors control it more precisely to align the robotic charging station and move it closer to the socket on the ship. Finally, the precise location information is transmitted to the PLC, and the robotic charging station transfers energy until the charging process is completed. Consequently, the proposed system achieved 94% accuracy in charging socket detection. This research can be a guide in demonstrating the shore-to-ship socket detection efficiency of autonomous and computer vision-based ship charging systems.