National Day Special II: Integrating physical hardware with intelligent "brain" technology, the 15th Five-Year Plan drives the humanoid robot industry onto a fast growth track

"14th Five-Year" plan period marks the first year of prototype verification for China's humanoid robot industry. The iteration speed of complete products has accelerated. Core hardware components such as servo motors, precision reducers, and dexterous end-effectors are increasingly being developed domestically. Intelligent algorithms like large-scale embodied adaptation, motion control, and environmental perception are continuously improved. Upstream and downstream innovation enterprises are emerging continuously. Technology research and development, prototype testing, and small-scale pilot applications are all accelerating. With the help of China's mature automotive electronics and precision manufacturing industry chain, product costs continue to decrease. The industry's development focus has shifted from early laboratory prototype demonstrations to small-batch delivery and application testing. Embodied large models further enhance the robot's environmental perception and task planning capabilities. However, at the same time, high-end components and high-precision control systems still have shortcomings. The self-reliant industrial chain has not fully formed yet, which does not match the demand for rapid mass production. Meanwhile, the number of industry pilot scenarios is expanding rapidly, but a standardized and replicable commercialization model has not yet matured. The profitability and sustainability of the industry need to be verified, and the overall industry is still in the initial stage of industrialization.

During the "15th Five-Year" plan period, humanoid robots will move away from a development model focused on small-scale technical demonstrations and enter a critical window period for batch supply of components and operational implementation in specific scenarios. As the maturity of complete products increases and the industry chain gradually improves, the focus of industry development will shift from simply enhancing motion control, dexterous operation, and intelligent interaction capabilities to solving problems such as mass production, reliability of core components, cost reduction, and adaptability to complex scenarios. Complete product manufacturers will face pressures such as mass delivery, product reliability, supply chain coordination, and cost reduction throughout the entire lifecycle. Application companies will need to address issues such as scenario adaptation, deployment costs, operation efficiency, and actual production efficiency improvements. In the future, it is necessary to coordinate humanoid robot bodies, core components, embodied intelligence, data training, application scenarios, and safety standards, and accelerate the establishment of a technology system covering perception, decision-making, execution, and continuous learning, promoting humanoid robots to transition from single-machine demonstrations and local pilots to stable, frequent, scaled, and sustainable commercial applications.

The "15th Five-Year" plan includes robots in the development system of strategic emerging industries and also lists embodied intelligence as a future industry for forward-looking layout, promoting deep integration of artificial intelligence with advanced manufacturing and robotics technology. The "15th Five-Year" plan makes systematic arrangements from multiple dimensions, including embodied intelligence training infrastructure, model algorithms, humanoid robot bodies and core components, and applied scenarios, aiming to enhance robot perception, decision-making, motion control, and human-robot interaction capabilities. Key components such as high-performance actuators, dexterous hands, and sensors are being developed. A training and testing platform adapted to embodied intelligence is being established. Humanoid robots are being applied in fields such as industrial manufacturing, logistics and warehousing, and commercial services, gradually achieving scale implementation. The focus of humanoid robot industry development has shifted from previous technical research and prototype iterations to optimizing the performance reliability of complete products, reducing production costs, improving the data and training system, and relying on real-world application to drive the entire industry towards scale development. We believe the following developments are particularly noteworthy:

In terms of industry positioning

The "15th Five-Year Plan Outline" clearly states "accelerate the development of the robot industry and promote embodied intelligence to become a new economic growth point," making top-level strategic arrangements for the humanoid robot industry. The "Notice on Jointly Carrying out the 2026 Humanoid Robot and Embodied Intelligence Practical Training Special Action" issued by the Ministry of Industry and Information Technology and the State-owned Assets Supervision and Administration Commission of the State Council proposes "by the end of 2026, create 100 high-value scenarios and achieve ten-thousand-unit-scale implementation capabilities." At the same time, the "Guidelines for the Construction of a Comprehensive Standardization System for Humanoid Robots and Embodied Intelligence" will be released soon, paving the way for large-scale implementation of humanoid robots. A series of policy "combination punches" will enable humanoid robots to move beyond technical research and small-scale demonstrations and quickly enter a new stage of development with policy support, large-scale scenario verification, and continuous improvement of the standard system.

Regarding core componentsside

The plan emphasizes accelerating the development of high-speed precision bearings, high-parameter gears and transmission devices, high-reliability hydraulic and pneumatic seals, high-performance motors and control systems, high-precision screws and other core basic components. Although the layout is aimed at advanced manufacturing as a whole, relevant categories include key components such as humanoid robot joint motors, transmission mechanisms, screws, and bearings, which will determine the long-term cost and reliability ceiling of domestic robot joints and reducers.

Regarding model collaboration

The plan proposes to coordinate the layout of embodied intelligence training facilities, promote virtual-real integrated collaborative training and evolution, and conduct in-depth research on physical artificial intelligence, developing embodied models and algorithms integrating the brain and body. Policies focus not only on robot hardware but also on the development progress of "brain and body" technologies, focusing on promoting the development of robot intelligents with decision-making and motion control capabilities, enabling humanoid robots to truly enter daily life scenarios.

Regarding infrastructure

The plan encourages the construction of a new generation of supercomputing, general computing, and intelligent computing facilities, actively develops public cloud services, builds a computing power monitoring and scheduling platform, and formulates and improves standards for computing power resource pooling, grid connection, monitoring, operation, and scheduling. Policies not only emphasize computing power outside, but also ensure the layout of domestic humanoid robot supporting bases, providing underlying support for large-scale robot simulation and data closed-loop training in the future.

Regarding application scenarios

The plan proposes implementing large-scale application demonstrations of new technologies, new products, and new scenarios, and increasing scenario cultivation and openness. Scenarios with clear needs and paying entities, such as special rescue, industrial manufacturing, urban governance, public services, and personal companionship, will become the main breakthrough points for humanoid robots to transition from demonstration applications to large-scale operations.

Domesticization of core components is accelerated, and the industry's development foundation is initially formed

During the "14th Five-Year" period, the market share of core robot components in China continued to increase. Taking six-dimensional force sensors as an example, from 2020 to 2025, the sales share of Chinese-funded six-dimensional force sensors increased significantly from 19% in 2020 to 59% in 2025, reversing the position in both domestic and foreign markets. Local manufacturers have grown from minor players to industry leaders, proving that domestic enterprises have completed product verification, mass implementation, and captured the mainstream market in the field of core robot sensing components.

On the other hand, data shows that after a leap in 2024, the domestic share only increased by 1% in 2025, indicating that foreign investors are retreating from the remaining high-end market. The remaining share is mostly concentrated in high-reliability and high-precision high-end industrial scenarios, with higher technical barriers and difficulty in rapid competition. Market substitution is shifting from "easy-to-achieve mid-to-low-end markets" to tackling high-end markets, making marginal substitution more difficult.

Source: Frost & Sullivan analysis

In the domestic robot harmonic reducer market, Japan's Hamner's market share decreased from 36% in 2021 to 32% in 2025, while the combined share of domestic manufacturers increased from 65% to 68%. Domestic harmonic reducers are continuously eroding Japan's market share, but the substitution speed is relatively slow. Domestication has entered a stage of "increasing volume but still lacking in high-end development".

Source: Frost & Sullivan analysis

During the "15th Five-Year" plan period, the robot component industry will focus on addressing high-end performance weaknesses such as product lifespan, accuracy, and stability. By opening up real application scenarios such as industrial and special operations, it will provide engineering verification platforms for domestic component companies, helping them accumulate scenario data. At the same time, policies will accelerate the establishment of a robot component performance and safety standard system, remove the certification threshold for high-end customers, and shift the industry's competitive focus to high-value-added high-end markets.

The overall market size of Chinese humanoid robots is expected to grow from approximately 5.1 billion yuan in 2025 to approximately 20.2 billion yuan in 2030, with an average annual compound growth rate of 31.7% during the "15th Five-Year" plan period. The growth engine of the market size mainly comes from the maturity of domesticization of core components and continuous optimization of algorithm models, which are expected to significantly reduce the cost of robot complete products to less than 100,000 yuan, thereby bringing a significant increase in product output.

The focus of the "15th Five-Year" plan is no longer limited to the production and manufacturing of humanoid robot complete products, but rather increasing support for high-speed precision bearings, high-parameter gears and transmission devices, high-reliability hydraulic and pneumatic seals, high-performance motors, control systems, high-precision screws, and sensing components, which are key core components. At the same time, embodied intelligence training is regarded as a key area for humanoid robots to achieve practical applications. This also indicates that the underlying hardware conditions required for humanoid robots to achieve joint movement and force sensing, as well as the actual operating capabilities of robots, will be significantly strengthened during the "15th Five-Year" period.

As the industry enters the stage of small-batch application scenario tests, the industry landscape also shifts to comprehensive competition. Enterprises must not only tackle core hardware, improve the reliability of joint movement and force sensing, rely on supply chain capabilities to reduce costs and achieve stable mass production, but also strengthen embodied intelligence training, accumulate real operation data, and improve the robot's environmental adaptability and practical operation capabilities. Whether it is possible to establish a closed loop of hardware, algorithm, and scenario, and run a feasible commercial return model will be the key to distinguishing enterprise competitiveness.

Source: Frost & Sullivan analysis

Implementation of large-scale application demonstrations of new technologies, new products, and new scenarios during the "15th Five-Year" plan period, increased scenario cultivation and openness, providing a platform for real-world testing and data accumulation for humanoid robots, accelerating product iteration and verification, and exploring feasible business models, play a key guiding role for the industry.

From the perspective of commercialization conditions, scenarios such as guest service and warehousing logistics have relatively low technical thresholds, and basic navigation and simple interaction functions are easy to implement, making them suitable for large-scale pilots in the early stage of industrialization. They are expected to be the demonstration scenarios prioritized for implementation during the "15th Five-Year" period. Industrial manufacturing, disaster relief, and other industrial and special scenarios have clear needs and large market spaces, but require high precision in motion control, complex working environments, and difficult technical implementation. They are still in the continuous engineering verification stage. Overall, the industry will prioritize scenarios with mature technology and calculable commercial returns to enter the market.

Extreme scenarios such as space operations and underwater operations have great difficulty in technical implementation, requiring extremely high environmental adaptability and safety reliability. Many technical challenges such as water pressure, vacuum, radiation, and autonomous decision-making also need to be overcome. During the "15th Five-Year" plan period, the focus will be on conducting software and hardware engineering verification, exploring business models, and conducting small-scale demonstration pilots in multiple scenarios. The core goal is to bring the motion control and operation capabilities of humanoid robots to a preliminary practical level, and such ultra-high-difficulty extreme scenarios will not be promoted at this stage.

Source: Frost & Sullivan analysis

Commercialization of the humanoid robot industry relies heavily on the decline in complete product costs. The main approaches focus on flattening fixed costs through large-scale mass production, domestic substitution of the supply chain, reduction of hardware redundancy combined with software algorithm optimization, and the transformation of the industry from non-standard parts to standard parts. During the prototype development stage, customized development, small-batch production, and purchase of expensive overseas components together increased the cost of complete products. After entering the mass production stage, the establishment of standardized production lines, local supply chain support, and software-hardware optimization, combined with the gradual implementation of industry unified standards, will significantly reduce the cost of complete products. At the same time, algorithm optimization can compensate for hardware performance weaknesses, reducing the industry's dependence on high-end hardware and creating space for the commercialization of humanoid robots.

During the "15th Five-Year" plan period, priority will be given to accelerating the maturity of the domestic supply chain, promoting domestic substitution of overseas components, and taking embodied intelligence training as a key area for development. Relying on large models and motion control algorithm optimization can achieve software-hardware cost reduction. Combined with the promotion of new technologies, new products, and new scenario demonstrations during the "15th Five-Year" plan period, demonstrating leading product solutions in multiple scenarios can help the industry form a unified technical standard, promote the transformation of components from non-standard to standard parts, continuously reduce the cost of complete products, and support the expansion of demonstrations in priority scenarios such as warehousing and logistics, industrial manufacturing.

Source: Frost & Sullivan analysis