Data and AI-driven reform now spans the entire auto operation process of industry giant China FAW Group Co., Ltd. This was made abundantly clear at the 9th Digital China Summit held in Fuzhou, southeast China's Fujian province in late April, when the company displayed its latest achievements in the digital and intelligent transformation of auto manufacturing.

In a traditional auto manufacturing system, a large amount of data is separated into isolated information silos. To deal with such sophisticated data issues, FAW developed an Enterprise Operation Agent (EOA).

Men Xin, assistant to the general manager of FAW and manager of the system digitalization department, said the EOA aims to build an intelligent operation hub that independently perceives, makes decisions, executes and evolves. The hub utilizes data, models and agents to optimize the enterprise's full value chain administration dynamically.

In the Fanrong plant area of the FAW Hongqi Manufacturing Center in Changchun, northeast China's Jilin province, orders for customized automobiles enter the production scheduling process according to production orders automatically generated by the system. Driven by data, production organization, process control and quality management operate in a highly efficient and coordinated manner.

The credit for such smooth operation should be given to "digital staff." Driven by the intelligent operation hub, manual approval steps have been reduced by 60 percent, the R&D and production cycles have been significantly shortened by 50 percent, and manufacturing costs have dropped by 40 percent.

The production plan administrator is the enterprise's "digital staff No.001," and there are 20 such "digital staff" members now. The enterprise has realized the intelligentization of 23 key decisions with the help of these "digital staff."

AI has also been integrated into the entire process of FAW's auto R&D and production. The company has compiled a methodology of digital and intelligent transformation that reorganizes the business logic via "data+AI."

On the R&D side, FAW developed GPT-Code, a code generation AI application that can assist tasks like functional testing, code commenting, and the automatic generation of front-end code from prototypes, cutting the coding time by 50 percent.

On the manufacturing side, FAW uses AI large models as the "brain" to develop process agents and production agents, realizing intelligent decision-making and automated production scheduling.

Virtual simulation technologies are used as the "brain" to build a cloud-controlled simulation technology platform, enabling decoupled hardware-software cluster control.

The brick-and-mortar factories are the "torso," accelerating the development of embodied intelligent robots to be deployed in scenarios such as final assembly and logistics sorting.

Digital and intelligent technologies have significantly boosted the company's operational efficiency. Men explained that by fully adopting cloud-native technologies, FAW migrated 290 standalone systems to the cloud in a single, coordinated effort, reducing business iteration cycles from quarterly to weekly and slashing annual maintenance costs from 200 million RMB to 27 million RMB.

FAW's transformation path provides a highly practical and replicable model for the upgrading of traditional industrial bases.