This study proposes a novel method to improve cargo container and port infrastructure inspections by integrating drone technology, edge computing and advanced machine learning. Drones equipped with high-resolution cameras and sensors capture real-time data, facilitating instant anomaly detection. Edge computing reduces latency, enhancing inspection efficiency. The YOLOv8n model precisely identifies structural anomalies like cracks and corrosion. Synthetic data generation diversifies the training dataset. XGBoost classifies corrosion severity, while convolutional neural networks measure crack lengths. This integration signifies a significant departure from manual methods, promising heightened efficiency, accuracy and reliability in maintaining crucial port infrastructure. The study underscores the potential of combining these technologies for real time, automated inspection processes, thus improving safety and efficiency in port operations.