The first of its kind nationwide! Ningbo-Zhoushan Port introduces “robotic dogs” to assist customs inspections, ushering in a new paradigm for intelligent inspection of heavy containers.

On March 2, at the heavy-container inspection area in the Meishan Port Zone of Ningbo-Zhoushan Port, a four-legged “robotic dog” nimbly maneuvered among the containers, precisely checking the container numbers and seals of the containers awaiting inspection—this marks the first time that an inspection-assisting robot has taken on the task of automatically capturing and identifying container numbers and seal numbers at ports nationwide. This initiative has pioneered the application of embodied intelligence on customs heavy-container inspection platforms across China’s ports, setting a new paradigm for intelligent collaboration between customs and ports and accelerating the transition of heavy-container inspection operations toward greater intelligence and efficiency.

Embodied intelligence empowers customs inspection, solving the “time lag” challenge.
         Embodied intelligence refers to intelligent systems that rely on physical entities to perceive, understand, and act. Through real-time interactions between embodied agents and their environment, these systems can autonomously acquire information, analyze problems, and execute tasks. Leveraging the development opportunities presented by embodied intelligence, Xintong Company, in collaboration with Maidong Company, keenly identified these emerging prospects, ventured deep into frontline port operations, and repeatedly refined and optimized their solutions. Building upon the "Chengfan" AI large model, they successfully developed an auxiliary robot system for container inspection. This system precisely addresses the "time lag" issue inherent in the collaborative operations between customs and terminals, driving a transformative upgrade in inspection methodologies—from the traditional approach of "on-site manual inspections and individual verification" to a more advanced model of "intelligent sensing and efficient approval."

Facing the challenge of surging throughput, intelligently breaking through efficiency bottlenecks.
         In recent years, as the container throughput at Meishan Port Area has continued to rise, the volume of inspection work has grown by more than 10% annually. Meanwhile, the capacity of the port’s inspection facilities has nearly reached saturation, and there is an increasingly acute shortage of human resources, both of which are hindering improvements in port operational efficiency. Under the existing inspection procedure, customs officers must individually photograph the container numbers and seal numbers, verify their accuracy, then open the container doors and take a full-image photograph of the cargo before finally initiating the re-stowing operation.

         Today, customs officers can issue operational instructions through the relevant system, while the inspection-assist robot autonomously plans the route and safely avoids obstacles to locate the target container. Equipped with a high-definition camera and AI recognition algorithms, the robot precisely captures information such as container numbers and seal numbers. Throughout the entire operation process, image records are simultaneously transmitted to customs, creating a complete, traceable, and verifiable regulatory chain. After customs officers verify the quality of the images and confirm their accuracy, terminal operators can then proceed to carry out the container-opening and re-stowing operations in coordination.

Visible efficiency gains: Completed in 20 minutes, with an identification accuracy rate exceeding 99%.
         “The robots have already verified all the container information in advance, reducing the unpacking time to just 20 minutes and significantly boosting the efficiency of unpacking and re-stowing operations!” said Zhang Wenzhi, an inspector from the Second Inspection Section of Meishan Customs. Data shows that the current accuracy rate of container number recognition by the inspection-assisting robots exceeds 99%. When three inspection-assisting robots work simultaneously, tasks that previously took four to six people more than an hour to complete can now be finished in under 20 minutes.

         The successful application of this “robotic dog” represents an innovative practice in which Zhejiang Seaport Group, Ningbo-Zhoushan Port Group, and Ningbo Customs have deepened their collaboration and promoted the development of a smart port. Moving forward, both sides will continue to deepen the application of intelligent inspection technologies, constantly refine the accuracy of AI recognition algorithms, improve the human-machine collaborative scheduling mechanism, and further expand the scope of embodied intelligence applications in more port scenarios. Together, they will strive to create a new industry benchmark for smart, collaborative customs-port operations, injecting stronger intellectual momentum into the construction of a world-class, powerful port.