The rapid advancements in deep learning technologies are transforming the way visual inspections are conducted in the railway industry. Traditional methods of monitoring railway vehicles by human inspector are both labor-intensive and costly. To address these challenges, this paper focuses on the application of deep learning, specifically convolutional neural network (CNN) for automated object detection undercarriages of railway vehicles in real-time. The integration artificial inteligence solutions into railway monitoring especially for visual inspection of railway vehicles promises significant cost reductions, improved operational reliability, and enhanced safety standards. This paper also presents a case study demonstrating the effectiveness of CNN (YOLOv5) in detecting and identifying critical components in the undercarriage of railway vehicles, showcasing the potential of deep learning in the future of railway visual inspections.