Recently, at the site of the renovation project of Tianjin Port Pacific International Container Terminal, the new generation of truck positioning system based on artificial intelligence vision independently developed by ZPMC has completed functional debugging. It marks the successful operation of the artificial intelligence vision subsystem under the first ZPMC intelligent crane architecture, which is an important step for ZPMC in the application of intelligent perception in smart ports and provides technical support for the subsequent intelligent identification scenarios.

The project is based on the fourth sub-topic of ZPMC's "Unveiling" project Truck Guidance and Positioning System CPS. The hardware structure of the system is simple and powerful, which can realize the simultaneous guidance function of multi-lane trucks. The accuracy of double-box detection and recognition is as high as 99%. At the same time, it has the functions of truck mobile identification and anti-smashing protection of the front of the car, providing stable and reliable safety protection for the terminal. The accuracy of the large and small car direction position detection accuracy of the system is within ±5cm, and the rotation angle detection accuracy is ±0.2°, which provides efficient and accurate positioning data for the automatic landside truck automation of the shore bridge.

The intelligent crane architecture is a distributed, modular system architecture based on Linux and artificial intelligence technology, and each module of the system operates independently and interacts with real-time data. The architecture standardizes the external communication interface and data format of the system, and each module of the system has a unified web browser for configuration and debugging, which can be quickly customized and developed to realize the integration of a stand-alone perception system.

With the on-site commissioning of Tianjin Port Pacific International Container Terminal, the project team has completed the data interaction and debugging of a new generation of the truck positioning system and ZPMC automation control system, which initially realizes the one-key loading function of landside truck automation in quay cranes, which is a verification of ZPMC's independent development of intelligent crane architecture concept. The adoption of this architecture is more conducive to the localization of alternative development of subsequent system-use equipment, improves development efficiency, and provides intelligent solutions for ZPMC's intelligent and unmanned terminal project.