1. National Strategic Expectations Compared to New Energy Vehicles
The four major expectations of the national strategy: The country has listed humanoid robots as a new track, with four strategic expectations. First, to promote technological innovation and breakthroughs. Humanoid robots integrate advanced technologies such as artificial intelligence, requiring breakthroughs in key technologies such as the brain (e.g., developing large AI models) and limbs (e.g., robotic arms, dexterous hands), to build a complete technological innovation system. Second, to cultivate new economic growth points. Humanoid robots are seen as a disruptive product following computers, with a market size expected to be comparable to electric vehicles by 2035. The country hopes to cultivate a new economic growth engine through this, injecting new momentum into the economy. Third, to promote industrial upgrading and transformation. Humanoid robots are widely used in various fields, driving the industry from labor-intensive to intelligent and efficient upgrades, improving production efficiency and quality, and assisting China in developing towards high-end industries and building a new industrial system. Fourth, to address population aging and labor shortages. Predictions show that by 2100, China’s population may drop to 700 million, creating significant population pressure. The development of humanoid robots involves changes in labor factors, which can alleviate structural contradictions in the labor market.
Comparison with the positioning of new energy vehicles: Humanoid robots share similarities and differences with new energy vehicles. Similarities: Both are strategic emerging industries, important forces for promoting economic and industrial upgrades, supported by national policies, and enhance China’s position in the global industrial chain through technological innovation and industrial development, driving the coordinated development of upstream and downstream industrial chains. New energy vehicles drive the development of the battery industry chain; humanoid robots drive the development of the upstream precision machinery, midstream intelligent chips, and downstream application production industry chains. Differences: First, in terms of technological complexity, humanoid robots involve multiple disciplines and require significant breakthroughs in various aspects; the technology of new energy vehicles focuses on batteries, etc. Second, in terms of application scenarios, humanoid robots are applied in various fields, having a more profound impact on human society; new energy vehicles are mainly used for transportation. Third, in terms of development stages and market maturity, new energy vehicles have matured after one or two decades of development; humanoid robots are in the early stages of development, with investments in the seed round. Although there has been progress, there are shortcomings in complete machines, and the current VRA model is considered a transitional solution, requiring increased R&D and market cultivation efforts.
2. Future Industrial Pilot Zone Layout and Policy Support
Key regional industrial clusters: The layout of future industrial pilot zones in the humanoid robot field is mainly concentrated in core areas such as the Yangtze River Delta, Beijing-Tianjin-Hebei, Pearl River Delta, Wuhan, and Hefei. The Yangtze River Delta, as an important industrial cluster, is represented by Zhangjiang in Shanghai and Suzhou: Under the promotion of Secretary Chen Jining, Zhangjiang has become a core carrying area for humanoid robots, gathering over 70 core enterprises such as the International Center for Humanoid Robots, Zhiyuan, Fourier, Kepler, Ant Lingbo, and Green Harmonics, with an expected overall industrial scale of 70 billion by 2025; Suzhou has a developed manufacturing industry, gathering several enterprises such as New Haitu, Leju, Galaxy General, and Ulrich, and the Yangtze River Delta is a major gathering area for Tesla’s industrial chain, including Top Group in Ningbo, Zhejiang (linear actuator assembly), Sanhua Intelligent Control in Shaoxing, Zhejiang (rotary joint assembly), Mingzhi Electric in Qingpu (hollow cup motor), Green Harmonics in Suzhou (harmonic reducer), Beite Technology in Jiading (precision ball screw), Wuzhou Xinchun in Shaoxing (planetary roller screw), Keli Sensor in Ningbo (six-dimensional torque sensor), Rongtai Co., Ltd. in Yangzhou (motor module), and Riying Electronics in Changzhou (electronic skin). The Beijing-Tianjin-Hebei region, centered on Beijing, has the Beijing Economic Development Zone (Yizhuang) with a hundred billion innovation fund supporting key technological breakthroughs and has established a national-local co-built humanoid robot innovation center (expected to be upgraded to national level by 2025). Tianjin and Hebei have a good industrial foundation and can serve as production and component supply bases. The Pearl River Delta region (Greater Bay Area) is a global supply chain gathering place for humanoid robots, with numerous supporting production enterprises in Shenzhen, Guangzhou, and Dongguan, where Nanshan has formed a complete industrial chain layout for complete machines, with leading enterprises such as UBTECH, Leju, and Pudu establishing their presence here, and Shenzhen has the Guangdong Humanoid Robot Innovation Center. Wuhan is focusing on humanoid robot complete machines with an academic team at its core, with Huawei occupying 50%-60% of the electronic skin (tactile sensor) market and having a hundred billion-level fund invested in the industrial chain. In Hefei, the provincial innovation platform, pilot verification platform, and investment incubation platform have been established at the Guoxuan Center; Wuhu plans to invest 1.2 billion to build the largest humanoid robot training ground in the country (the Beijing Shijingshan training ground only invested 130 million). Additionally, Mianyang in Sichuan may form a western industrial cluster based on Changhong and Tianlian.
Policy and resource support measures: The policy and resource support for humanoid robots in future industrial pilot zones mainly cover funding, subsidies, innovation platforms, talent, industrial exchanges, and scene opening. In terms of funding support, local governments have established hundred billion-level industrial support funds; regarding project subsidies, Pudong is brewing a 30% subsidy and computing power vouchers for humanoid robot enterprises. In terms of innovation platform construction, the Hubei Humanoid Robot Innovation Center has been completed by 2025, providing common technology services and promoting collaborative innovation in the industrial chain; the Yizhuang Innovation Center in Beijing has been established (expected to be upgraded to national level). In terms of talent support, local governments are inclined towards the robot field in national or provincial talent cultivation, such as the Shanghai Innovation Center significantly supporting the number of pearl talents and leading talents compared to other industries. In terms of industrial exchanges and cooperation, Shanghai hosts the Humanoid Intelligence Conference, Developer Conference, and WAISA Artificial Intelligence Conference, while Beijing hosts the World Robot Conference, and Hangzhou hosts the iROSE Conference to promote industrial exchanges. In terms of application scene opening, local governments organize local central state-owned enterprises and robot companies to connect, opening scenes for enterprises to collect data and test, promoting application landing.
3. Industrial Development Planning and Path During the 14th Five-Year Plan
Policy guidance and special planning: In terms of policy connection, the humanoid robot development guidance opinion will be released in November 2023 during the 14th Five-Year Plan, and more special plans or guidance opinions are expected to be issued during the 15th Five-Year Plan to refine the industrial development path. In terms of policy guidance, the Ministry of Industry and Information Technology (MIIT) has clarified that the 15th Five-Year Plan will focus on the frontier fields of humanoid robots, opening up new tracks, building pilot zones, and formulating special paths for various fields, with a focus on breakthroughs in dynamic balance control, simulation perception, and other technologies, promoting their large-scale application in industrial, service, and special scenarios. In terms of planning outline, the 15th Five-Year Plan is expected to include macro expressions such as the industrialization of intelligent agents and the construction of general platforms. The industrial strategy shifts from “technological breakthroughs” in the 14th Five-Year Plan to “system construction and commercialization” in the 15th Five-Year Plan, with the goal shifting from “breaking through key technologies” to “building global competitiveness,” reflecting the differences in the different development stages of the industry.
Specific implementation measures: In terms of technological breakthroughs and platform construction, led by the MIIT and the Ministry of Science and Technology, a national innovation center is jointly built with universities and enterprises, with more than ten provincial-level centers already established. Through the “接榜挂帅” (taking the lead) approach to tackle key technologies, testing and certification centers are established, such as the first Chinese-style base led by Hangzhou Wubai Intelligent, which has invested 200 million. In terms of special projects and standardization promotion, drawing on the experiences of integrated circuits and new energy vehicles, it is proposed to establish the “Humanoid Robot 2030 Project” to support project R&D and enterprise breakthroughs. In addition, the MIIT plans to establish a standardization technical committee to lead innovation, standardize development, and strive for the definition of international standards.
4. Bottlenecks in Mass Production and Commercial Application
Technical and scenario challenges: Currently, the maturity of domestic humanoid robot technology is significantly insufficient. In terms of motion control, the poor heat dissipation of motors can lead to temperatures reaching 60-70°C during outdoor operation in summer, resulting in increased power consumption, decreased joint stability, and a higher risk of falling. The embodied intelligence technology has not converged, with the mainstream VOA algorithm being a transitional solution, and the future technical form is unclear; the world model is still in the research stage and is expected to mature in 10-20 years; currently relying on imitation learning, generalization capabilities are achieved through a large amount of data combined with the Transformer model. Expanding application scenarios is difficult, with unclear commercialization paths, mostly for performance displays, and the value of replacing human labor is not clear. The cost of data collection is high, with a single data point being about 2GB, and the Great Wall task requiring hundreds of GB of storage; the cost of a single Leju robot is 500,000 yuan, with low data output efficiency and a long cost recovery cycle. 2025 is seen as the year of mass production for domestic humanoid robots based on Tesla’s original plan to produce 5,000 units, but the actual output is less than 100 units, making this definition no longer valid.
Safety, cost, and acceptance: The commercialization of humanoid robots faces challenges in safety, cost, and user acceptance. In terms of safety, there are risks in human-robot close interactions, and the responsibility for robot-related injuries in home scenarios is unclear. In terms of cost, product prices range from 100,000 to 1 million yuan, while the price of data collection robots has decreased (from about 300,000 at the beginning of the year to below 100,000 in July, and below 70,000 this month), but overall remains high; the cost of a single Tesla Optimus A5 chip is 5,000 USD, combined with the costs of the complete machine and model training, achieving a target price of 20,000 to 30,000 USD is challenging. In terms of user acceptance and regulations, robots are mostly used in training grounds or specific closed areas, requiring a user education process similar to that of new energy vehicles in their early stages; supporting regulations need to be improved, such as insurance and annual inspection mechanisms not being clearly defined. Although 2025 or 2026 may be seen as the year of mass production, there is still a long way to go from small-scale production to large-scale commercial application.
5. Development Expectations and Quantitative Indicators for 2030
Industry scale and localization goals: In terms of industry scale, the market size is expected to be about 20 billion in 2025, increasing to 350 billion by 2030, with an average annual compound growth rate of about 70%. The localization rate of core components is required to reach over 90%.
Enterprise and application scale: The goal for enterprise cultivation is to nurture at least 5 enterprises with a market value of 100 billion, and over 100 specialized and innovative small giant enterprises. The application scale is expected to reach millions, benchmarking Tesla’s production target of 1 million units by 2030. In terms of technological innovation, about 100,000 invention patents are expected. In terms of standard systems, plans are to formulate more than ten international standards for humanoid robots and establish CNF-recognized laboratories.
6. Technical Route Guidance and Competitive Landscape
Technical route policy guidance: The policy adopts a “combination of guidance and market selection” approach, avoiding excessive intervention, and guiding resources through pilot projects, innovation consortia, and standard system construction. Pilot projects include “robot + intelligent manufacturing” in the automotive and electronics industries and smart elderly care communities, providing support through scenario cultivation. Innovation consortia encourage collaborative breakthroughs in industry-university-research cooperation, and the MIIT may organize breakthroughs through the “接榜挂帅” approach and set goals for advancement. Policies support the localization of core components such as high-performance servo motors and precision reducers, clarifying key areas for technological breakthroughs and improving the localization rate of high-precision reducers. In 2025, the CPPCC will hold multiple forums on the silver economy, with the central government focusing on elderly care, but robot technology is not mature, and rehabilitation and elderly care robots are important directions for the next 5-10 years, with short-term effects being difficult to see.
Competitive landscape and ecological cooperation: Leading enterprises in the domestic humanoid robot market have significant advantages, with Yushu and Zhiyuan performing outstandingly. Yushu stands out for its extreme mobility, self-developed motors, and reducers to reduce costs, with sales expected to grow 4-5 times compared to last year in 2025, with many applications in schools and other scenarios. Zhiyuan relies on the Huawei ecosystem, laying out the entire industrial chain, attracting developers by opening middleware, and supply chain companies take pride in cooperating with it. UBTECH’s Walker S2 series has received nearly 600 million in intended orders for 2025, with some deliveries by the end of the year, leveraging the automotive market layout. Leading enterprises pose challenges to new entrants in terms of technology, funding, market share, and brand recognition. However, technology has not converged, and new entrants have opportunities for leapfrogging, as evidenced by Datar’s closure in 2024 after receiving over 5 billion in financing. Ecological cooperation is becoming mainstream, with Zhiyuan improving its ecological system, and Huawei collaborating with UBTECH and Leju to launch empowerment plans to support small and medium-sized enterprises. The industry shows characteristics of “the new pushing the old” iteration, with limited long-term survival for leading enterprises.
Q&A
Q: In the national industrial strategy, the 15th plan lists humanoid robots as a new track. What expectations does it carry? How does it compare to the positioning of new energy vehicles?
A: The national strategic expectations for humanoid robots mainly include four aspects: promoting technological innovation and breakthroughs; cultivating new economic growth points; promoting industrial upgrading and transformation; and addressing population aging and labor shortages. The similarities with the positioning of new energy vehicles are that both are regarded as strategic emerging industries, supported by policies, enhancing global industrial chain status through technological innovation and industrial development, and driving the coordinated development of upstream and downstream industrial chains. The differences are that humanoid robots have higher technological complexity, broader application scenarios, and are in an earlier development stage.
Q: In recent years, the number of graduates has reached tens of millions. The automation of ride-hailing and taxi services, along with the penetration of humanoid robots, may impact professions such as drivers. How should this be balanced in the short term?
A: Short-term balance is mainly related to the lag of the education system. Currently, it takes about 5 years to cultivate students with new technical capabilities in colleges, vocational schools, and undergraduate institutions. Existing graduates have received training from the old education system, leading to a mismatch between skills and industrial upgrade needs, resulting in limited career choices, with many turning to ride-hailing drivers and other positions. This phenomenon is a short-term pain. The country is promoting the emergence of new job opportunities related to humanoid robots, such as data collectors, maintenance personnel, and secondary development personnel, through industrial support, allowing existing graduates to re-enter education to access new career opportunities.
Q: During the 15th Five-Year Plan, how will the national level formulate and refine special industrial development plans or guidance opinions for the robot field?
A: In terms of policy background, the humanoid robot development guidance opinion was released in November 2023, and more special development plans or guidance opinions for the humanoid robot field are expected to be issued during the 15th Five-Year Plan to refine development paths. In terms of policy guidance, the MIIT has clarified that the 15th Five-Year Plan will focus on the frontier fields of humanoid robots, systematically opening up new industrial tracks, building future industrial pilot zones, and formulating special development paths for various fields, with a focus on breakthroughs in dynamic balance control, simulation perception, and other key technologies, promoting large-scale applications of humanoid robots in industrial, service, and special scenarios. The planning outline is expected to include macro expressions such as the industrialization of intelligent agents, the construction of general humanoid robot platforms, and basic research plans for humanoid intelligent agents. The national strategy will shift from research support to system construction and commercialization orientation, with the 15th Five-Year Plan for robot industry development expected to be a derivative version of the 14th Five-Year Plan, with strategies shifting from breaking through key technologies to building global competitiveness. Specific measures include: the MIIT and the Ministry of Science and Technology leading the joint construction of a national humanoid robot innovation center with universities and enterprises; tackling key technologies through the “接榜挂帅” approach, establishing standard testing systems and product certification centers; implementing key special projects; and the MIIT planning to convene innovation centers, complete machine and component enterprises to establish a humanoid robot standardization technical committee to lead technology innovation, standardize industry development, and strive for the definition of international standards.
Q: What is the current progress and status of the standard system construction for humanoid robots in China?
A: Previously, there was no standard construction for humanoid robots in China, but traditional automation research institutions such as the Institute of Automation and the Institute of Control have standard-setting experience and are actively participating. Currently, the standard-setting entities are scattered, and although they attempt to formulate standards based on new energy vehicle standards, the standards that have been issued lack responsiveness, and their enforceability is questionable. The standard system includes levels such as group standards, local standards, industry standards, national standards, and international standards, but the construction process lacks a dedicated institution for overall planning, leading to reverse promotion issues. Recently, the MIIT held a meeting to plan the establishment of a humanoid robot standardization technical committee, led by the humanoid robot hundred-person committee under the China Electronics Society, aiming to build a systematic standard system blueprint.
Q: Is the construction of the humanoid robot standard system still in its early stages, and has it recently been promoted by higher levels?
A: The construction of the humanoid robot standard system has been emphasized by various regions since early last year. Due to the emerging nature of the industry, enterprises or institutions that formulate standards have the motivation to profit from standards, but there are issues with standards not being recognized. Currently, standard construction is relatively scattered, lacking a dedicated department for overall planning. Recently, the MIIT held a meeting to propose the establishment of a humanoid robot standard technical committee, led by the humanoid robot hundred-person committee under the China Electronics Society, marking the beginning of high-level promotion.
Q: Does the construction of the standard system for the domestic machinery industry need to be led by a higher level from the top down?
A: Yes, it is needed. Currently, the humanoid robot standard system is being constructed by the humanoid spirit hundred-person committee under the China Electronics Society, which is an important progress.
Q: How to assess the current distance from small-scale production to large-scale commercial application of domestic humanoid robots? What are the main bottlenecks?
A: 2025 is seen as the year of mass production for domestic humanoid robots mainly based on Tesla’s judgment, but due to Tesla’s production progress not meeting expectations, this definition has become invalid. Currently, domestic humanoid robots are still in the early stage of transitioning from pilot applications to large-scale promotion, with a long distance to large-scale commercial applications. The industrial field may require 3-5 years to achieve large-scale applications, while the household field will take longer. The main bottlenecks include: insufficient technological maturity; unclear application scenario demands; high data collection costs and low collection efficiency; safety and reliability needing verification; high costs and prices; and user acceptance and regulations needing improvement.
Q: According to the 15th plan, what will the humanoid robot development look like in 2030, and what are the quantitative indicators?
A: The MIIT proposed that the core goal for 2027 is to achieve a shipment volume of 100,000 units, forming 50 benchmark applications and 5-10 high-level gathering areas. By 2030, the quantitative indicators for humanoid robot development include: the industry scale increasing from the current approximately 20 billion to 350 billion, with an average annual compound growth rate of about 70%; the localization rate of core components reaching over 90%; nurturing 5 enterprises with a market value of 100 billion, and over 100 specialized and innovative small giant enterprises; achieving an application scale of millions; in terms of technological innovation, about 100,000 invention patents; and in terms of standard systems, formulating more than ten international standards and establishing CNF-recognized laboratories.
Q: During the 15th period, will the policy support for different technical routes in the robot industry remain neutral, or will it guide resources towards specific technical routes through pilot projects?
A: During the 15th period, the policy support for technical routes in the robot industry will adopt a combination of policy guidance and market selection, neither completely neutral nor overly intervening, mainly guiding resources towards routes with better industrialization prospects through pilot projects, innovation consortia, and other mechanisms. The policy will clearly support the localization of core components such as high-performance servo motors and precision reducers, making them key areas for technological breakthroughs; at the same time, it will provide application scenario support for different technical routes through large-scale application demonstrations in manufacturing and livelihood sectors. In addition, the policy will encourage innovation consortia to collaborate on breakthroughs, promote key technological breakthroughs through MIIT-led projects, and strengthen standard system construction guidance to clarify technical routes. It is worth noting that the central government places great importance on the silver economy, with rehabilitation and elderly care being important development directions for robots in the next 5-10 years, but due to the current technological maturity, significant effects are difficult to see in the short term.
Q: The leading enterprises in the domestic humanoid robot market are mainly Yushu, Zhiyuan, and others. In the current high market concentration, is the opportunity window for new entrants shrinking? Will ecological cooperation become the mainstream model?
A: Leading enterprises in the domestic humanoid robot market have significant advantages, but the opportunity window for new entrants has not completely shrunk, and ecological cooperation may become the mainstream model. The challenges posed by leading enterprises to new entrants mainly lie in two aspects: one is technological and funding advantages; the other is market share and brand recognition. New entrants have opportunities as current technology is not mature, and there is a possibility of leapfrogging through new technologies. Ecological cooperation is becoming the mainstream model, driven by the need for collaborative industrial development; leading enterprises are actively laying out ecosystems, and new entrants can gain development opportunities by participating in ecological cooperation.