China’s Robotics Revenue Soars as Industry Races to Crack Embodied AI
Listen to the full version
Revenue from China’s robotics industry surged 29.5% year-over-year during the first three quarters of 2025, outpacing expectations as the country produced 595,000 industrial robots and 13.5 million service robots — both exceeding the full-year totals for 2024.
The country’s density of manufacturing robots — measured as the number of robots per 10,000 employees — more than doubled since 2020, rising to 567 units and propelling China from eighth to third in the global ranking, according to Wang Weiming, a senior official at the Ministry of Industry and Information Technology. Wang disclosed the data during the 2025 China Robot Industry Development Conference held on Nov. 11.
The figures reflect rapid growth in a sector now shifting its focus to a more complex challenge: embodied intelligence. At a separate industry forum, investors and executives weighed the path forward as the race intensifies to integrate artificial intelligence with physical capabilities.
Unlock exclusive discounts with a Caixin group subscription — ideal for teams and organizations.
Subscribe to both Caixin Global and The Wall Street Journal — for the price of one.
- RELATED
- DIGEST HUB
Digest Hub
Back
- China's robotics industry revenue rose 29.5% year-over-year in Q1–Q3 2025, producing 595,000 industrial and 13.5 million service robots, surpassing 2024 totals.
- Robot density in manufacturing doubled since 2020 to 567 per 10,000 employees, moving China to third globally.
- The sector faces challenges in embodied intelligence, lacking unified frameworks and data, while firms push commercialization and ecosystem development.
- Alibaba Group
- Jiang Shanshan, head of frontier tech investment at Alibaba Group, discussed the challenges in embodied intelligence for robotics. She noted that current Vision-Language-Action (VLA) models are not optimized for robotics and lose crucial information. Jiang emphasized the need for companies to achieve phased profitability or secure financing to remain competitive.
- Lingchu Intelligence
- Wang Qibin is the CEO of Lingchu Intelligence. He highlighted three key challenges in the embodied intelligence sector: the lack of a unified model framework, limited real-world training data, and the absence of a broad user market to provide data feedback for rapid iteration.
- Nvidia Corp.
- Nvidia Corp. is actively developing "world models," a type of AI that produces physically accurate simulations of real-world environments. These models are seen as a foundational technology for spatial intelligence and are being explored by tech giants as the robotics industry shifts its focus to embodied intelligence.
- Alphabet Inc.'s Google is actively developing "world models," a type of AI that produces physically accurate simulations of real-world environments. This technology is considered a foundation for spatial intelligence and is seen as a possible path forward for the embodied intelligence sector in robotics.
- RoboScience
- RoboScience is a company whose founder and CEO is Tian Ye. The company emphasizes the urgent need for commercial application in the rapidly evolving field of embodied intelligence. They prioritize building an "ecosystem flywheel" to enable fast deployment across new use cases, highlighting the importance of generating revenue streams even as core technologies are still under development for this "marathon" industry.
- Ant Group
- Ant Group is mentioned in the article as an investor in the robotics industry. Han Chenglong, an investment director at Ant Group, suggests that startups in the robotics sector need to identify their niche to remain competitive, especially as leading firms move towards initial public offerings.
- Amio Robotics
- Liu Fang, the CEO of Amio Robotics, believes that while early profits in the robotics industry might go to component manufacturers, long-term success will belong to companies that provide comprehensive, end-to-end solutions and services. This suggests Amio Robotics may be focused on developing complete robotic systems rather than just individual parts.
- PODCAST
- MOST POPULAR
Sign in with Google
Sign in with Facebook
Sign in with 财新