Autonomous Robotic System for Visual Inspection in Electrical Substations and Cable Galleries

This article presents a research project focused on improving the efficiency and safety of maintenance processes in hazardous and difficult-to-access environments, such as openair electrical substations and cable galleries. A mobile robot integrates perception technologies to gather and process an inner representation of the environment allowing it to recognize known inspection targets. The developed prototype incorporates teleoperation and autonomous mission capabilities, as well as object recovery and incremental learning techniques to enhance detection accuracy. The article also details creation of both real and synthetic datasets for training and testing, as well as the evaluation of the robot's performance in object detection and tracking process. Results suggest that the developed technology can improve maintenance processes, reduce downtime and costs associated with equipment failures, and minimize occupational hazards.