A train is subjected to a comprehensive inspection using a camera gate. As the train passes through the camera gate, various images are taken from different angles by a total of 32 cameras. The data collected is processed by an AI that detects over 200 types of defects such as material wear or stone chip damage. Any deviations from the target condition detected are displayed to the technicians in the maintenance hall. A human checks the images to determine whether there is actually a defect. If a defect is confirmed, a work order is automatically generated and sent to the workshop staff's tablet.