China’s embodied artificial intelligence industry is experiencing explosive growth, with the market expected to surpass 1 trillion yuan (approximately $146.54 billion) by 2035. This projection, provided by the Development Research Center of the State Council, was a focal point at the recent World Intelligence Expo in Tianjin, as reported by Xinhua. The expo featured a dedicated “Robot Town” spanning 1,700 square meters and hosted over 700 exhibitors, offering a comprehensive snapshot of where China’s robotics and AI integration stands today.
International observers were struck by the pace of progress on display. Andrea Rösinger, a representative from a German technology company, offered a candid assessment: “My feeling is that China is leading the industry in robotics and making rapid progress in AI.” The observation reflects a growing consensus among foreign industry participants that China has moved from fast follower to genuine leader in the embodied AI space.
(Related: China Launches First Humanoid Robot Lifecycle Management Platform)
Manufacturing Depth as a Competitive Advantage
A recurring theme at the expo was the role of China’s industrial base in enabling rapid embodied AI development. Galileo (Tianjin) Technology Co., Ltd., a manufacturer of quadruped robots, has accumulated over 300 patents and software copyrights and has deployed its robots in demanding environments including rail transit systems, power grids, and explosion-proof petrochemical facilities. Yu Tong, the company’s international sales manager, articulated the underlying advantage: “The rapid development of embodied AI would not be possible without China’s solid, full-chain manufacturing base.”
This sentiment was echoed across the expo floor. Unlike software-only AI development, embodied AI requires tight integration between advanced algorithms and physical hardware, motors, sensors, actuators, and structural components. China’s dominance in manufacturing supply chains means that domestic robotics companies can iterate on hardware designs at a speed and cost that is difficult to replicate elsewhere.
PaXini Technology demonstrated the scale of China’s data collection infrastructure for embodied AI. The company recently released a 10-billion-scale multimodal embodied AI dataset, gathered through five super data-collection factories spread across China. Technicians at these facilities wear tactile gloves equipped with 30 six-dimensional tactile modules, capturing nuanced physical interaction data across 15 industrial and consumer sectors. Xu Jincheng, founder and CEO of PaXini, also announced the completion of a landmark cross-border transaction of embodied AI training data under Tianjin’s pioneering negative list policy for data outbound management, a regulatory first that could open new pathways for international data collaboration.
Open-Source Momentum and National Strategy
The open-source movement is gaining momentum in China’s embodied AI sector. AutoNavi has fully open-sourced its ABot-M0 platform, providing the broader research community with access to robot datasets, model architecture, training frameworks, and tool chains. This move mirrors the open-source strategies that have accelerated progress in large language models and positions China’s robotics ecosystem to benefit from distributed global contributions.
The Chinese government is actively reinforcing this momentum. Earlier this month, China launched a national-level AI pilot-testing base for embodied intelligence in Hangzhou, providing a dedicated facility for companies to test and validate their systems in controlled real-world conditions. Wu Qingxiang, an associate professor at Nankai University’s College of Artificial Intelligence, emphasized the collaborative ethos driving the sector: “China’s AI industry has always pursued an open, co-creative ecosystem.”
(Related: China’s Six Little Dragons of Robotics Sprint Toward IPOs Amid Funding Surge)
With China holding 60% of the world’s AI patents and its core AI industry exceeding 1.2 trillion yuan, the country is well-positioned to maintain its leadership in the development and deployment of embodied AI technologies. The 1 trillion yuan market projection for 2035 represents a roughly seven-fold increase from current levels, a trajectory that, if sustained, would make embodied AI one of the defining industrial transformations of the next decade.
The international dimension of China’s ambitions for embodied AI is also becoming clearer. PaXini’s cross-border training data transaction under Tianjin’s negative list policy is a small but significant step toward enabling Chinese embodied AI companies to operate globally. As humanoid robots and quadruped platforms move from research labs to commercial deployment, the ability to collect and share training data across borders will become increasingly important, and China’s willingness to create regulatory pathways for this activity signals that it intends to compete for global market share, not just domestic dominance. For international robotics companies, the message from the Tianjin expo is clear: China’s embodied AI ecosystem is no longer a domestic story.