A nationwide reskilling push aims to prepare industrial workers for automation, but the pace of change is testing both people and institutions.
By Fan Yiying and Chen Yongyi
GUANGDONG, South China — In a 200-square-meter workshop in the southern manufacturing hub of Foshan, 38-year-old Chen Zhen deftly adjusts the parameters on a robotic arm.
As the machine rotates beside him, a simulation mirrors each movement in real time — a tool the mid-career electrical engineer hadn’t operated independently until a specialized course last year.
“Now we just run the simulation on a computer instead of troubleshooting on site,” he says.
In Foshan and across China’s factory belt, he’s become exactly the kind of worker the national retraining system is trying to produce: upskilled, ambitious, and still standing in an industry rapidly shedding older roles.
Chen now manages full robot integration projects — work that once required specialists — at a company that helps manufacturers deploy and maintain automated production lines.
“Automation is evolving fast,” he tells Sixth Tone. “If we don’t keep up, we’ll be eliminated.”
That pressure is most visible in places like Foshan. Known for producing microwaves, air conditioners, and other consumer goods at scale, the city logged over 3 trillion yuan ($418 billion) in industrial output last year.
An industrial robot at German robotics giant KUKA’s workshop, Foshan, Guangdong province, 2022. Wang Qingwu/VCG
With nearly 80% of local companies expected to adopt some form of automation by year’s end, factories are upgrading machines and rewriting workflows — often faster than their workers can adapt.
At the same time, factories are struggling to find workers. China’s labor pool is shrinking as birth rates decline, and younger workers increasingly reject traditional factory jobs in favor of e-commerce or gig work.
To close the gap, Chinese authorities have launched an ambitious retraining effort, where local governments, schools, and companies are experimenting with joint programs to help workers catch up.
In Guangdong alone, the provincial government has pledged 5 billion yuan over three years to reskill millions of industrial workers, offering subsidies, certification incentives, and customized corporate training schemes — all aimed at building a workforce fluent in automation before the machines get too far ahead.
So far, the results are uneven. Some workers, like Chen, find a new foothold. Others are pushed sideways — or out entirely. And as automation advances, a future on the factory floor comes down to two options: adapt or exit.
Robotics system accessories undergo testing at a company in Foshan, Guangdong province, June 2025. Guo Jun/CNS/VCG
On the line
In Foshan, Chen has seen that binary choice unfold up close.
At his company, which handles the operation and maintenance of robotic production lines, manufacturers regularly send workers for retraining after new automation systems are introduced.
Some return with new roles. But the lines never look the same. “The overall workforce on a given production line always shrinks by about 15%,” he explains. “When it comes to industrial robots, I think a lot of repetitive jobs will inevitably disappear.”
As a systems electrical engineer, Chen says he’s felt the pace of change “intensely.” Some colleagues are already struggling to keep up with the demands brought on by AI and robotics. “There’s just not much time to retrain — especially mid-career, when you’ve got family and everything else,” he says.
In Foshan alone, digital upgrades have delivered a 16% rise in productivity and cut costs by 17%. And across Guangdong, annual robot output has climbed from 44,700 units in 2019 to 246,800 this year — a fivefold increase that has kept the province the top producer of industrial robots in China for five consecutive years.
At Wingotech Technology in the tech hub of Shenzhen, an AI vision system now inspects 400 to 600 sealing rings per minute — a task that once required 10 to 20 workers using magnifying glasses. It now runs with just two or three operators.
“Think of our system as 32 inspectors working in sync,” says Qin Lin, general manager of Wingotech. “In aerospace, a single faulty seal can cause catastrophic failure.”
Qin Lin inspects an AI system in Foshan, Guangdong province, June 2025. Fan Yiying/Sixth Tone
Qin displays a gasket ring sample (left) alongside his company’s AI system, Foshan, Guangdong province, June 2025. Fan Yiying/Sixth Tone
The system uses high-resolution cameras and AI to automate everything from part feeding to defect detection and sorting. One of its production units now doubles as a training platform at Shunde Polytechnic University.
Operators require only basic software training, though the real challenge, Qin says, is adapting the system for thousands of seal variations. “Some workers are reluctant to learn,” he says. “But many seize the opportunity to improve — moving from basic roles to operators, and even managers.”
Across the industry, low-skilled roles are becoming harder to protect. “Packaging is one of the first to go,” says Fan Qiyuan, associate professor of industrial robot technology at Shunde Polytechnic University. “But at the same time, the demand for people who can manage, maintain, or improve these systems is growing.”
“Tomorrow’s packaging worker won’t box items by hand,” Fan asserts. “They’ll oversee and optimize robotic packaging lines instead.”
Chen agrees. “Before upskilling, I shared those concerns,” he says. “Now with proper training, I view robotics as an opportunity rather than a threat.”
Associate professor Fan Qiyuan demonstrates smart factory functions supported by robotics systems, Guangdong province, June 2025. Fan Yiying/Sixth Tone
Learning to last
At Career Education in Foshan — a private training center focused on manufacturing technical training — instructors say the hardest part is correcting the assumptions trainees bring with them.
“They think operating robots is just about pushing buttons,” says Lao Youqian, a veteran instructor who has taught at the center for five years.
Career Education launched its industrial robotics program in 2015 and has since trained more than 1,000 students through two tracks: government-subsidized corporate partnerships and self-funded individual courses.
According to Lao, electrical engineers show exceptional engagement, seeking to master programmable logic controllers and robotics. But for absolute beginners, the conceptual shift proves challenging.
“To address this, we often invite technical supervisors from other companies to share career growth paths, showing them the progression from operator to engineer,” says Lao.
Instructor Lao Youqian adjusts training equipment at Career Education, Foshan, Guangdong province, June 2025. Fan Yiying/Sixth Tone
One such case is Jiang Jing, an employee at Canbo Electrical Appliances, who funded his own training. He moved from assembly line worker to technical manager, and now teaches part time at Career Education.
“Students value Jiang’s courses because he brings real-world factory experience,” says Fu Xiaoguang, principal of Career Education. “Company-sponsored trainees often lack motivation, whereas career-driven individuals show remarkable progress.”
The institution’s faculty includes four full-time instructors, supported by a dozen vocational teachers and industry experts. While enrollment in standalone robotics courses remains modest, integrated automation programs — combining robotics with other technologies — attract stronger demand, especially from Foshan-based manufacturers sending employees for upskilling.
Yet demand doesn’t always translate to motivation. Lao says trainees sent to Career Education often treat the course as a formality. “When it’s mandatory, they show up, but they don’t engage,” he says. Linking training to salary increases or recognition, he adds, could help shift that mindset.
A robotics system in use at Career Education, Foshan, Guangdong province, June 2025. Fan Yiying/Sixth Tone
“Multinationals and state-owned enterprises invest in workforce development,” says Fu. “But with private firms, we have to prove the value first.” Some offer overtime pay for training hours, but most small- and mid-sized companies still treat it as a cost, not a priority.
Fu believes that needs to change. He points to Germany’s tax incentive model as one China could adopt to push companies toward broader participation.
“China is at that point now,” he says. “We should prioritize automation-related upskilling for current employees to minimize layoffs.”
Scaling the gap
In Guangdong, that warning is already shaping policy. Under the “Million Talents Plan,” the province has pledged to reskill 3 million industrial workers over three years, with a focus on AI-related roles. The goal is to keep frontline staff employable as automation spreads.
Domestic media refer to this as tackling the “replacement-creation paradox,” where machines erase some jobs and open others, often faster than workers can adapt.
Foshan alone hosts dozens of industrial robotics training programs, including one at Shunde Polytechnic University, launched with government backing in 2022. The school partners with top manufacturers like Midea and offers certification subsidies tied to hands-on courses to align technical training with real factory needs.
The tripartite model — schools, government, and enterprises — helps spread costs: companies receive subsidies when employees earn intermediate or advanced robotics certifications.
Robotics systems inside the training lab at Shunde Polytechnic University, Guangdong province, June 2025. Fan Yiying/Sixth Tone
Robotics systems inside the training lab at Shunde Polytechnic University, Guangdong province, June 2025. Fan Yiying/Sixth Tone
Shunde Polytechnic University’s training lab spans 200 square meters and houses 11 robotics systems from major brands like KUKA and ABB. It doubles as both a certification site and a teaching space designed to reflect real factory environments.
“If students can master advanced skills like visual detection and 3D positioning,” says Fan, “they won’t be replaced — they’ll expand their career paths.”
Zhou Qiang, 35, is banking on that. He left civil engineering for industrial robotics and is now training to operate commonly used systems. “I realized I couldn’t stay in civil engineering indefinitely,” he says. “At my age, changing careers is still possible — and my family fully supports the move.”
And Qu Rui, 24, made a similar choice at the start of his career, turning down a construction job at a state-owned firm to enroll in robotics training instead. “The civil engineering market is oversaturated,” he says. “And between the unpaid overtime and unclear future, I’d rather start lower in a field that’s growing.”
But even for workers who adapt, the bar is rising.
At Zhinanche — a training firm based in the eastern tech hub of Hangzhou that specializes in field-ready industrial robotic automation engineers — instructors say employer expectations have shifted noticeably in recent years.
“Where a technical diploma once sufficed for maintenance roles,” says Li Xiaoling, who teaches there, “employers now consistently require college degrees as minimum qualifications.”
Li Hui has been with Aisan, a Japanese-owned auto parts plant in Foshan, for 13 years. In the past two years, he’s had to complete two rounds of robotics training and now holds an advanced certificate. His colleague Kang Jianbo, a die-casting specialist, puts it plainly: “Technical jobs are still secure, but mastering these new systems is now mandatory.”
Li Hui at work. Courtesy of Li Hui
The speed of change is testing the limits of the training ecosystem itself.
At Shunde Polytechnic University, Associate Professor Deng Xia says the school is struggling to scale up its outreach. “We actively recruit enterprises, but many still lack incentives to prioritize upskilling,” she says. With just 20 full-time faculty and administrative staff, most balancing teaching with other responsibilities, the program is stretched thin.
Factory realities add to the difficulty. “Smaller firms operate with leaner teams and lower automation levels,” says Associate Professor Fan, adding that irregular work schedules, including shift rotations and frequent business travel, make consistent attendance difficult. “Training during work hours disrupts production, as each worker typically mans a single station.”
Training centers too face mounting pressure to upgrade both equipment and curricula as robotics tools evolve. “Just adding collaborative robots or AI vision systems can cost over a million yuan,” says Career Education’s Lao. “And instructors need regular upskilling, too.”
The impact of automation won’t be immediate, says Xu Jianhua, associate professor at Peking University. But the risk is clearest at the bottom. “Low-skilled manual jobs in manufacturing are among those most at risk,” she says. “However, vocational training programs address both high-tech and low-tech occupations.”
Fan agrees. “Robotics is inevitable,” he says. “The only real choice is whether companies and governments help workers keep up.”
Liu Zenglong, founder of Zhinanche, sees potential even in transition. Most of his trainees still come from mechanical, electrical, or computer science backgrounds, but interest is rapidly widening.
“They may not become inventors,” he adds, “but they can develop into highly capable applied professionals who solve real-world automation challenges.”
Additional reporting: Isaac McDonald; editor: Apurva.
(Header images: A robot inside the training lab at Shunde Polytechnic University, Guangdong province, June 2025. Fan Yiying/Sixth Tone)