The industry you once laughed at is about to create 1 million new jobs
Amid the wave of embodied intelligence, humanoid robots—previously dismissed by many as merely "good at dancing"—are rapidly moving into reality, driving a surge in emerging occupations.
The first to take off is the role of data collector, i.e., people who specialize in "teaching robots how to work." In factories simulating real-world scenarios, they wear sensing devices and perform daily actions—folding clothes, opening bottle caps, making oatmeal—while their movements are synced in real time to robots.
During this process, data such as joint angles and motion trajectories of the robot are collected. To ensure data validity, movements must be precise, smooth, and completed within a specified time, repeated hundreds of times daily. After standardization, this data is supplied to R&D and manufacturing enterprises for training robots to autonomously perform tasks.
According to Zhilian Zhaopin's "2026 China Embodied Intelligence Robot Industry Development Talent Report," data collector job postings in 2025 increased by 769% year-on-year, while data annotation/AI trainer positions surged by 2250%.
Combining data from Liepin's "2026 Robot Industry Talent Supply and Demand Trend Insight Report" and corporate recruitment information, current data collectors with technical experience or on-site roles earn a monthly salary of 8,000–15,000 yuan on average, while senior or management-level positions can reach over 20,000 yuan. Although lower than algorithm engineers (starting at 30,000 yuan), this role has low entry barriers, high demand, fast salary growth, and can transition to higher-level positions such as data annotation management or robot trainer with experience.
This is just a glimpse.
According to CCTV Finance, China's embodied intelligence industry is projected to reach a scale of 400 billion yuan by 2030 and exceed one trillion yuan by 2035, driving over 1 million jobs.
What does 1 million mean? It is equivalent to the annual new employment target of a major economic province like Guangdong or Jiangsu.
In short, humanoid robots are no longer just "iron lumps" that dance on exhibition stands; they are building a new set of industrial and living infrastructure from scratch. These 1 million jobs represent a brand-new track emerging from zero, roughly divided into three categories:
First, the "body-building" hardware camp: precision engineers working on reducers and servo motors, system engineers assembling parts into reliable products, and production process engineers. This is an upgraded version of manufacturing, requiring a large number of high-end technicians and engineers.
Second, the "brain-training" software camp: not writing ordinary apps but developing embodied intelligence algorithms and simulation testing, letting robots first learn to walk and work in a virtual world before moving to reality. Algorithm engineers and simulation modelers are the core.
Third, the "service-providing" deployment camp: integration engineers designing automation solutions for factories, operation and maintenance engineers responsible for repairs, and "robot coaches" teaching workers how to use them. This does not include the supporting supply chain, quality inspection, sales, and after-sales service.
Looking at recent recruitment big data, humanoid robots/embodied intelligence are currently in a "talent-grabbing and team-building phase" transitioning from lab R&D to industrial mass production, not yet in the large-scale assembly line worker recruitment stage, but core R&D and data positions have begun to explode, and the humanoid robot track is becoming a new high ground for high-tech employment and salaries.
According to the "Maimai 2026 Spring Recruitment Workplace Insight Report" released by Maimai Gaopin Talent Think Tank, from January to April 2026, the recruitment index in the embodied intelligence field reached 579, a 15-fold increase from 36 in the same period of 2025. According to Liepin Big Data Research Institute's "2026 Robot Industry Talent Supply and Demand Trend Insight Report," new job postings for humanoid robots increased by 215.8% year-on-year in the past year, with an average annual salary of 406,100 yuan for related positions.
According to reports from Yicai and other media, the monthly salary of an embodied intelligence algorithm engineer generally starts at 30,000 yuan, with senior positions reaching 50,000–80,000 yuan, and some large-model architect positions listing up to 80,000–110,000 yuan.
Regionally, according to Liepin, Shanghai, Shenzhen, and Beijing together account for over 40% of new job postings, but new first-tier cities such as Wuhan (+110.28%), Hefei (+104.23%), and Suzhou (+103.74%) have year-on-year growth rates exceeding 100%.
Analysis suggests that humanoid robots and embodied intelligence have entered a positive cycle of "technological breakthrough → capital influx → talent competition → cluster formation," and positions in core components, vertical scenario applications, and data services still have significant growth potential.
Unfortunately, despite the talent gap of 1 million people, Chinese universities currently have no corresponding majors, and the training system lags behind the industrial explosion.
The good news is that the Ministry of Education recently released the "Undergraduate Programs Catalogue for General Higher Education Institutions (2026)," adding an undergraduate major in embodied intelligence at nine universities, with curriculum and ability training closely aligned with the industry's demand for compound talents, promoting the adaptation of the university training system.
Now, if you are already in the workforce and have some knowledge of hardware, algorithms, or data processing, it is best to keep an eye on policy hubs like Beijing, Shanghai, Shenzhen, Hangzhou, Suzhou, and Hefei, and choose a core-tech startup as an "early employee," betting on the dividends when the industry explodes in the next three years.
The trillion-level reservoir has already been opened.
Author: Ma Jiangbo
