Summary of Key Points
The recent boom in the field of embodied intelligence (robots that have bodies, can perceive, and can act like humans) has been incredibly intense: Yuzhu Technology has become the “first stock in the humanoid robot sector,” with its shares soaring by more than six times on the first day of trading; Meikamand, which provides “eyes, brain, and hands” for robots, is planning to list on the Hong Kong Stock Exchange; and Lingxin Qiaoshou, a supplier of robotic dexterous hands, has a valuation of 20 billion yuan. Capital is no longer solely focused on companies that manufacture complete robots; instead, it is pouring into those that supply core components, “brains,” and data for these robots. In the first half of the year, domestic embodied intelligence received 93.5 billion yuan in funding, with over half of that going to these upstream players. However, being an “upstream supplier” is not a guaranteed profit-making venture: manufacturers of complete robots are developing their own core components, and cross-industry players are entering the market, posing the risk of overcapacity. To succeed in the long term, suppliers must transform from being “replaceable” to “irreplaceable” partners.
Why Are Everyone Selling Components to Robots?
The core logic behind selling components to robots is that manufacturers have not yet found a stable path to profitability, but upstream components are essential for all robots:
- High uncertainty among manufacturers: Currently, robots either can only perform simple tasks (such as dancing or boxing) with human control or fail to meet production targets (like Tesla’s Optimus). No manufacturer can claim long-term profitability. However, components like joints, dexterous hands, and “brains” are indispensable for all robots—just as shovels are needed in a gold rush.
- Upstream components account for a large portion of costs: For example, Yuzhu’s G1 robot has joint modules accounting for 66% of the total cost, and sensors for 13.7%. As robot production scales up, the demand for these components will increase.
- Commercialization drives upstream innovation: Previously, robots just needed to be able to move and perform basic actions; now, the market requires them to perform practical tasks (such as moving items in factories or packaging in retail). This has led to the upgrading of upstream components.
Which Components Are the Most in Demand?
Capital is particularly interested in components that determine a robot’s “intelligence” and “dexterity”:
1. Embodied brains (the “brains” of robots): These are crucial as they determine whether a robot can understand its environment and make decisions on its own. For instance, Yuzhu’s founder mentioned that large robot models represent the biggest bottleneck for widespread application. Companies investing in this area, such as Delta Intelligence, which raised $300 million in three rounds within three months in January 2026, and Xingdong Jiyuan, which has a valuation of over ten billion yuan due to its advanced “brain” technology, are attracting significant investment.
2. Dexterous hands: These are essential for robots to perform tasks effectively. A dexterous hand consists of dozens of micro-motors and hundreds of cables and requires precise manual assembly, making them very expensive (high-performance models can cost hundreds of thousands of yuan each). The demand for such components is growing rapidly, with sales expected to increase by 265% in 2026, and they are being used in industries like automotive and pharmaceuticals.
3. Data infrastructure: Robots need real-world data to learn how to perform tasks (such as grasping and manipulating objects). Since this data is not available online, robots must collect it in real-world scenarios. Twenty-five data companies raised 17 billion yuan in the first half of 2026; those with this data will have a competitive advantage in making robots more intelligent.
4. Joint modules: These account for the highest proportion of the total cost (66%) and are key to a robot’s mobility. Yuzhu develops its own joint modules, highlighting their importance. Companies like Haoneng are investing heavily in expanding their capacity for producing these components.
Risks in the Component-Selling Business
The current success of component sellers is not guaranteed, and there are three major challenges:
1. Manufacturers developing their own components: Manufacturers are trying to reduce their dependence on suppliers. For example, Yuzhu develops 90% of its components in-house, and only buys “iron parts” (such as joint shells) from external suppliers. Other companies, like Zhiyuan, have separated their dexterous hand business to compete directly with suppliers. Tesla’s Optimus V3 has seen a significant increase in the number of movable joints (from 11 to 22), further compressing the market space for suppliers.
2. Cross-industry players entering the market: Companies from non-robot industries are bringing new technologies and cost advantages. For instance, a water quality testing company has applied micro-spectral chips to robot sensors, using cameras to detect contact deformations and AI to calculate pressure, disrupting the traditional pricing structure.
3. Potential overcapacity: Capital is driving rapid expansion in production capacity. For example, Haoneng is building a production line for 5 million减速 gears, and Midea is planning to produce 300,000 units. This is similar to the overcapacity situation in the battery industry; components that are in high demand today may become surplus in the future, leading to reduced profits. For example, Yuzhu’s robot prices dropped by 72%, but its gross margin increased due to cost savings from in-house development. However, if manufacturers start price wars, suppliers will be the first to suffer.
The Future of the Industry
To succeed in the embodied intelligence sector, suppliers must move from being “replaceable” to “irreplaceable” partners. This requires:
- High technical barriers: Only companies that can produce dexterous hands efficiently and stably will have bargaining power (similar to CATL’s position in the battery industry).
- Wide cross-industry applications: Components should be useful in various industries, providing greater market stability.
- Deep integration with manufacturers: Suppliers should provide customized solutions that help manufacturers solve problems and reduce costs, making them indispensable.
In summary, the component-selling business is very popular, but it’s not a guaranteed profit. Suppliers need to become essential players in the supply chain to maintain their competitiveness in the long term.