When Mr Sun Huihai first began working at a factory in the southern manufacturing belt of Guangdong some 13 years ago, his colleagues were all humans.

Now, they are joined by more than 200 robots that can work around the clock, seven days a week, to help produce air-conditioners for home appliances giant Midea.

Rows of bright orange robot arms whir at all hours of the day, fishing freshly pressed plastic parts out of hot metal moulds and onto a long conveyor belt.

Driverless robots with blinking lights store these parts in a multi-storey warehouse, and later take them to be assembled into units that are sold in China and around the world.

The number of robots put to work on the factory floor increases every year, said Mr Sun, 37, who heads the plant’s engineering department.

“Every day, we think about how to upgrade and make manufacturing here more intelligent,” he told The Straits Times.

Scenes like this have become more common across China, as the “factory of the world” turns to robotics to sustain and turbocharge its manufacturing juggernaut.

Over the past decade, the number of industrial robots on China’s factory floors has increased more than six times to over 1.7 million, as companies grappled with rising wages and a shortage of workers willing to staff production lines.

China now has the world’s third-highest density of robots in its manufacturing industry, trailing South Korea and Singapore in first and second place respectively, according to the International Federation of Robotics’ figures for 2023, the latest available.

Their deployment is poised to increase further as China continues its transition from low-value, labour-intensive production to advanced manufacturing – a national priority.

“At any given time, China cannot do without the manufacturing industry,” said Chinese President Xi Jinping in 2023. “The state will strongly support the development of high-end manufacturing.”

Policymakers in China, wary of the hollowing out of industries which can occur when countries get richer, have long pushed for greater automation to keep factories competitive.

A decade ago, the government rolled out “Made in China 2025”, a plan to upgrade manufacturing and become a production hub for high-tech sectors such as robots.

Rebates, subsidies and other incentives have been offered to encourage factories to automate. A rise in domestic production of industrial robots has also reduced prices, making the machines more affordable.

Factories in China pumped out nearly 370,000 of such robots in the first half of 2025, up 35.6 per cent from the previous year, according to figures from the National Bureau of Statistics.

At the Midea factory in Nansha, Guangzhou, where Mr Sun works, there are 204 robotic arms and 82 automated guided vehicles. They are supplied by Kuka, a German industrial robot giant which the Chinese company bought over.

One section of the plant, where plastic parts for the air-conditioner are moulded and retrieved, is dubbed a “dark (heideng)” area. It is so named because of the high degree of automation: In theory, it can run without humans or any lights on, but in practice, it is brightly lit here at the plant.

Not every part of the factory is as automated, a costly endeavour. Humans are needed to staff assembly lines, maintain the machines, and check the quality of manufactured parts. The facility employs some 4,000 workers during peak season, Mr Sun said.

Elsewhere, other manufacturers of electrical items, electronics and cars – the main users of industrial robots in China – have also ramped up the use of technology on their factory floors.

“Dark factories” have become a buzzword to describe the most advanced of China’s production facilities. Such operations have reportedly been adopted by companies ranging from home appliance giants Xiaomi and Gree to automakers Changan and Zeekr.

As robot adoption picks up pace, one question that arises is: What will happen to the more than 100 million workers whom China’s manufacturing sector employs?

The automation drive has at times been dubbed “replace humans with robots (jiqi huanren)”. In 2021, Gree’s chairman said that the company’s “dark factory” had slashed the need for workers at the plant from 10,000 to 1,000.

In Mr Sun’s telling, employment at Midea’s air-conditioner factory has remained roughly unchanged from a decade ago. What has changed, he said, is productivity. The number of air-conditioners the factory produces has more than tripled from 2015, company figures show.

Academics Nicole Wu and Sun Zhongwei, who interviewed and surveyed factory workers in southern China just prior to the Covid-19 pandemic, found that these individuals were not too concerned about robots just yet.

“Contrary to the more pessimistic assessments of automation, most manufacturing workers in Guangdong – who are buffered by steady increases in demand and a chronic labour shortage – appear to be unfazed by technological change at present,” they wrote in a paper published in 2025.

As China’s birth rate falls and the population grows more educated, it has become more difficult for factory bosses to fill jobs, said Professor Sun Zhongwei, who studies industrial relations and social security at the South China Normal University.

He is not worried that the automation drive will go so far as to undermine the manufacturing jobs often seen as a means of stabilising employment, because market forces are at play.

Automation is a rational process, and industrial robots are a sizeable investment, Prof Sun said. “Companies will need to calculate whether the cost of the machinery justifies the wages saved.”

Still, he added, the biggest losers as manufacturing goes high-tech are lower-educated, older migrant workers who lack the skills to remain relevant. Many will have to return to their rural homes to do odd jobs, while others might find employment as service staff.

Back at the Midea factory, Mr Wang Liangcai, 26, an engineer, believes that his job is safe from automation for now.

“Equipment still needs to be maintained, it can’t do so itself,” he said. “But if you think about the long run... we also don’t know how things will be.”

- In association with ANN