How Hapag-Lloyd improved schedule reliability with ML-powered vessel schedule predictions using Amazon SageMaker

Hapag-Lloyd, a leading global shipping company, developed an innovative machine learning solution using Amazon SageMaker to improve vessel schedule predictions and reliability.

• Created a multi-step ML solution with four specialized models to predict vessel arrival times
• Integrated complex data sources including vessel schedules, AIS tracking, and port performance metrics
• Used XGBoost algorithm to develop models for Ocean to Port, Port to Port, Berth Time, and Combined predictions
• Implemented a hybrid inference solution with daily batch processing and real-time API capabilities
• Achieved 12% improvement in ETA prediction accuracy compared to previous statistical methods

The ML-powered solution helps Hapag-Lloyd enhance schedule reliability, reduce manual intervention, and provide more accurate arrival time predictions across their global shipping network.