Railways are an important part of public transportation; hence they need a better monitoring system to ensure passenger safety. This article discusses how a railway management system utilizes Convolutional Neural Networks (CNN) to classify and monitor the safety of passengers as they travel. The proposed approach employs deep learning and real-time video monitoring to detect potentially harmful acts such as joining trains without authorization, crowding on railway platforms, falling from railway platforms, and other suspicious behavior. Monitoring without human assistance. The area of using machine learning to handle safety events at rail stations is all about developing smart systems that can predict, halt, and reduce accidents. By using machine learning algorithms and techniques such as predictive modeling, anomaly detection, and natural language processing. These systems can analyze a large amount of data from many sources, such as security cameras, sensors, and event reports, to identify potential safety issues and notify authorities so they may act before anything terrible occurs. This reduces the likelihood of accidents occurring and makes the experience safer for both passengers and employees. The proposed Enhanced Convolutional Neural Networks (eCNN) model utilizes camera frames to generate regularization criteria that assess whether a passenger is safe or harmful. Preprocessing photos, extracting characteristics, and categorizing them are all automated methods for monitoring passenger safety on trains. This automated equipment increases railway safety by enabling authorities to monitor the train in real time and react quickly to any safety concerns that emerge. This may improve passenger safety and operational efficiency.