AI stowage system launched at Ningbo-Zhoushan Port

The Chuanshan Port Area of Ningbo-Zhoushan Port. [Photo/Tide News]

Ningbo-Zhoushan Port has unveiled a new AI-powered stowage system at its Chuanshan Port Area, cutting container loading planning time from hours to just 20 minutes.

Traditionally, stowage planning required operators to manually place and adjust containers — an effort that could take over three hours per vessel. Now, with a single command on screen, the new system generates a loading plan for 3,000 units in just 20 minutes.

The transformation is backed by 20 years of port operational data, integrated into the port's self-developed n-TOS system and used to train a large AI model. Chen Jiachao, algorithm engineer, and his team spent weeks collaborating with veteran stowage staff to capture expertise and embed it into the AI's decision making.

Container stowage involves over 10 key variables — cargo type, box weight, crane scheduling, ship structure, and more. The AI model quantifies metrics like container rehandling rate for plans that replicate the judgment of seasoned operators.

For loading 3,000 units, the system reduces unnecessary container reshuffling by over 30 moves per plan on average and saved more than 1,000 minutes in crane operation time. With stable deployment underway, the port estimates the system will save millions of yuan annually.