China Bets on AI and Robots to Protect Its Factory Edge

Forget bland AI-generated music, pragmatic China is using robots and AI to focus on the world of factories.

The nation is showing intelligence and rapidly deploying AI and robotics across factories and ports.

The idea is to keep its status as “the world’s factory floor” as costs rise at home and foreign governments, including a possible second Trump administration, threaten more tariffs on Chinese goods.

Dark factories

An article in the Wall Street Journal highlighted some useful stats and information about China.

The country installed 295,000 industrial robots last year—nearly nine times as many as the US and more than the rest of the world combined—pushing it past two million operational robots nationwide.

William Blake spoke of “dark satanic mills,” but at Baosteel’s “dark factory” in Shanghai, it’s a positive phrase as operators monitor production from control rooms as AI systems reduce the need for manual intervention. A deputy director told state media that human operators now step in only once every 30 minutes instead of every three minutes.

Officials see such upgrades as vital as China’s population shrinks, young people shun factory work and manufacturing wages climb above those in countries such as India.

The “factory brain”

In the central city of Jingzhou, home-appliance giant Midea runs its washing-machine plant with an AI “factory brain,” a centralized system that coordinates robots and machines on the shop floor.

Nearly a decade after acquiring German robotics maker Kuka, Midea now uses Kuka robots under this AI control system, which manages 14 virtual agents that coordinate production tasks. Humanoid robots carry molded components for inspection by 3-D cameras, while AI decides how to correct defects.

For human staff, AI-powered glasses flag common errors, and the company says some processes that once took 15 minutes now take 30 seconds. Midea reports its revenue per employee rose nearly 40% between 2015 and 2024.

In fashion, Bosideng, founded by down-jacket tycoon Gao Dekang, has partnered with Zhejiang University to build an in-house AI model for design. The company says AI-generated concepts and virtual garments have cut the time to produce a clothing sample to 27 days from 100 and reduced development costs by 60%.

Huawei and heavy industry

Chinese officials acknowledge that in cutting-edge chips and frontier AI models, the country still trails the US. Yet firms such as Huawei and the model developer DeepSeek are central to Beijing’s strategy of catching up and deploying what exists at scale.

Huawei’s Pangu family of large language models and industrial AI services are being embedded directly into factories. At Conch Group’s giant cement operation in Wuhu, Huawei engineers have helped build tools to predict clinker strength and control kiln energy use.

Workers now watch as an AI model adjusts production in real time, while miles of conveyor belts are monitored by AI to speed up responses to breakdowns. Conch says it can now predict clinker strength with more than 85% accuracy, up from 70% with manual estimates, and that AI has already cut coal consumption by 1% on one production line—worth about $300,000 a year—with a 2% reduction targeted by the end of 2026.

The sea! The sea!

Ports are a special priority for Xi Jinping as China seeks to cement its dominance in global trade. At Tianjin, one of China’s biggest ports, operators working with Huawei have deployed unmanned trucks and a planning system called OptVerse AI Solver to optimize ship arrivals, crane capacity and other variables.

Planning work that used to take 24 hours now takes 10 minutes, according to Huawei. The port has also introduced PortGPT, an AI model that analyzes video and images on site and could eventually replace some human safety inspectors.

State media say more than 88% of large container equipment at Tianjin is already automated and that its highly automated operations require 60% fewer workers than traditional ports. For visitors, a promotional video proclaims: “We are the future.”

By contrast, only one of 10 large US container ports surveyed by the Government Accountability Office had deployed driverless vehicles by mid-2023, and just five were using AI or machine learning. US dockworker unions have pressed to restrict fully automated terminals and the use of AI for clerical tasks in recent contracts.

Competition and risks

China’s rush into AI-powered manufacturing comes as Donald Trump campaigns on pledges to bring back US industrial jobs through tariffs on Chinese goods and as Washington tightens export controls on advanced chips.

Chinese leaders are betting that automation will let them offset labor shortages and maintain cost advantages even as wages rise and demographics worsen. One survey cited by officials found that 83% of Chinese respondents believe AI-powered products and services are more beneficial than harmful—around double the share in the US—making rapid deployment politically easier.

Still, the experience at factories such as Conch shows AI rollouts are often slow, requiring time-consuming trial and error. While China is moving quickly to automate everything from washing machines to cement kilns to port logistics, the full impact on jobs, productivity and global manufacturing competition will likely unfold gradually.

UBTECH recently began delivering hundreds of its Walker S2 units across China, the first large-scale industrial deployment of full-size humanoids anywhere in the world.