In 2025, the scale of humanoid robots is on the rise. Previously, Figure AI and Ubiquitous robot Walker S1 entered factories to "work" and began exploring the industrial commercial use of humanoid robots. Tesla Optimus set a target of producing 5000 units annually by 2025 as early as the beginning of the year. And car companies such as Xiaopeng, Xiaomi, and Huawei will also release more substantial actions in the field of humanoid robots.

In the first half of 2025, with the promotion of technology, capital, scenarios, policies, and standards, the industrial ecology of humanoid robots will accelerate. Most opinions believe that in 2025, humanoid robots will embark on a new journey of mass production and commercialization. The industry may usher in its first explosive cycle in the second half of 2025.

But as the core power component of humanoid robots, battery companies had previously appeared calm. On the one hand, robot companies are already sending deeper cooperation signals to battery companies, but currently, the collaboration between the two parties is still in the early stages of exploration; On the other hand, battery companies have previously maintained a certain wait-and-see attitude after weighing technology investment and market returns.

It is understood that the collaboration between the two sides was once "unsatisfactory". Industry experts have pointed out that upon closer inspection of the battery pack of humanoid robots, the manufacturing level of their power system is still the same as the cylindrical battery cells and corresponding powertrain methods used in drones and robots in the past. This simple 'transplant' is difficult to continue meeting the increasing power requirements of future humanoid robots.

However, it is not surprising that the collaboration between the two sides has encountered setbacks. Humanoid robot companies have previously stated that neither the single machine installed capacity nor the installed scale of humanoid robots can be compared to the business of battery companies in the new energy vehicle sector. Meanwhile, the ideal power battery for humanoid robots still requires high performance and cost, and the current technical conditions for power batteries are not yet met. Collaborative deep development is not a small investment for battery companies.  

This points out the practical considerations behind battery companies' wait-and-see attitude: strict technical requirements and unfair economic accounts. The result of multiple factors such as mismatch between technology and demand, market size, and cost constraints is that the collaboration space between the two parties is currently limited.

As the popularity of humanoid robots continues to rise, battery companies with technological strength choose to take the lead at this stage, exploring dynamic balance solutions during the high-speed growth period of humanoid robots, and laying out the supply chain for humanoid robots in the first year of mass production.

At the same time, as new technologies such as 46 series large cylinders, semi-solid state batteries, and solid-state batteries continue to mature, these new technologies will also become a good opportunity for battery companies to compete in the field of humanoid robots.

Strict technical requirements and unbalanced economic accounts

are the power requirements for humanoid robots?

If it is the ultimate goal of replacing human labor, the requirements for power batteries will approach the optimal level in terms of energy density, rate, safety, wide temperature range, and cost.

For example, the Uber Walker S, which works in the Extreme Krypton factory, performs tasks such as collaborative sorting, collaborative handling, and precision assembly, which correspond to the three major characteristics of robots that can replace human labor: repetitive labor, physical labor, and high-precision work. So maintaining the operation of humanoid robots requires at least two basic requirements: high energy density and high rate capability of the battery.

In terms of energy density, battery packs are usually integrated into the chest cavity of humanoid robots. When the entire body exceeds 70kg, the weight reserved for the battery is usually 5-6kg. In small spaces and small weights, the energy density of the battery needs to be as high as possible to achieve the goal of replacing human labor for 8 hours a day.

But in reality, short battery life is a common problem for humanoid robots at present. The Tesla Optimu uses a battery capacity of 2.3 kWh and a range of approximately 1.5-2 hours. However, the charging time takes 0.5-1 hour. So if it is a usage scenario that requires long-term labor, two or more robots need to be equipped for rotation, resulting in greater cost expenditures.

In addition, energy density is not only related to endurance, but humanoid robots are currently exploring modular power supply systems that decompose centralized power into multiple collaborative subsystems. For example, the Yushu Go2 uses small capacity batteries (about 50Wh) specifically for the control system to ensure stable operation of core functions; Large capacity batteries provide power for the sports system. This also means that batteries require higher energy density to achieve a more compact and flexible module layout.

In terms of magnification, stable discharge capability is also a key factor in supporting the work of humanoid robots, especially in industrial grade scenarios that replace human physical labor for handling.

In terms of environmental adaptability, this involves two major performance indicators: safety and wide temperature range. In the working environment of humanoid robots, high-frequency charging and discharging cycles are a major safety hazard. Battery thermal runaway not only damages the robot body, but also endangers the surrounding environment.

And wide temperature tolerance is aimed at enabling humanoid robots to penetrate a wider range of scenarios without being affected by extreme heat and cold environments. But in fact, the problem of cold affecting the range of power batteries has not been completely solved, which also affects the penetration of new energy vehicles in the northern market.

Last but not least, cost is crucial. With the goal of replacing human labor, humanoid robots have been extremely sensitive to battery costs since the beginning of industrialization. As embodied intelligent agents, batteries account for less than 1% of the cost of humanoid robot BOM. The Tesla Optimus uses a 2.3KWh battery pack, with a cost of 2180 yuan, accounting for 0.5% of the total cost.

A humanoid robot company has stated that they can accept a maximum of 5000 yuan per package. They have high expectations for the next generation of high-performance battery technology, but the cost of using current solutions such as semi-solid state batteries and lithium metal batteries will skyrocket. At the same time, the cycle life and rate performance cannot meet the requirements.

Within the limitations of multiple considerations, humanoid robots can only find the optimal solution within the current level of lithium battery technology. At present, the main solution adopted by humanoid robots is the 18/21 series small cylindrical battery, with an energy density of 240-300Wh/kg and a rate capability of 3-5C. With the current level of material and process design in the industry, it can balance safety and wide temperature performance.

For example, the main battery supplier of Yushu Robot currently is Azure Lithium Battery, and its battery solution mainly uses cylindrical product system.

The importance of battery companies is increasing

Solid state batteries may become a new opportunity for cooperation

GGII estimates that the market size of Chinese humanoid robots will reach 2.158 billion yuan in 2024 and nearly 38 billion yuan in 2030, with a compound annual growth rate of more than 61%. The annual sales of Chinese humanoid robots will grow from about 4000 to 271200 units. By 2035, the sales of To B scenarios will reach 755000 units, with a market size of approximately 75.5 billion yuan; To C scenario sales will reach 1.256 million units, with a market size of approximately 62.8 billion yuan.

In terms of battery demand, GGII expects the demand in the robotics field to reach 2GWh by 2025 and 100GWh by 2030.

The extremely high growth rate is difficult for battery companies to ignore, especially with the rapid industrialization achievements since the beginning of the year and the outbreak of the humanoid robot market. Multiple battery companies have disclosed their layout in the field of humanoid robots.

As mentioned earlier, the "transplantation" of battery solutions in fields such as drones is difficult to meet the power needs of future humanoid robots. This also means opportunities for power battery companies. With the continuous maturity of new technologies such as 46 series large cylinders, semi-solid state batteries, and solid-state batteries, these new technologies will also become good opportunities for battery companies to enter the field of humanoid robots. Especially many companies focus on using solid-state battery technology to enter the field of humanoid robots.  

At present, semi-solid state batteries are a relatively mature product in the industry, with landing cases in emerging application markets such as drones and eVTOLs.

In terms of cost, the relevant calculations at the end of 2024 indicate that the cost of semi-solid state batteries considering a yield rate of 90% is 0.87 yuan/Wh, and the expected level of mainstream battery cost for new energy vehicles (0.3 yuan/Wh) is around 2035. But with a cost of 0.87 yuan/Wh, there is hope to open up space in the field of humanoid robots in the next stage.

Xinwangda has entered the field of humanoid robots through the polymer solid electrolyte route, and the solid-state battery it provides has been installed in the GoMate project and is being promoted for mass production.

At the 2025 Global Partner Summit, Honeycomb Energy officially proposed that "eVOLT+humanoid robots" will become its emerging business focus. The Honeycomb Plan divides solid-state batteries into semi-solid and all solid phases: the first generation of semi-solid batteries has an energy density of up to 300Wh/kg and is expected to be mass-produced by 2025; The second generation semi-solid state has a capacity of 360Wh/kg and 78Ah, and is designed for low altitude aircraft positioning; The energy density of the first generation all solid state battery will exceed 400Wh/kg and 68Ah, which will be applied in low altitude flight and automobiles, and also has the potential to provide energy for high-performance humanoid robots.

Funeng Technology has been deeply involved in soft pack batteries and solid-state batteries for many years, laying the foundation for it to seize opportunities in the humanoid robot market. Funeng Technology has launched three generations of products in the field of semi-solid state batteries, all of which have made orderly progress. The second generation of semi-solid state batteries adopts a new oxide/polymer solid electrolyte coating and densification technology, with an energy density of up to 330Wh/kg, a fast charging capacity of over 3C, a cycle life of over 4000 cycles, and a 90% energy retention rate at -20 ℃; The energy density of the third-generation semi-solid state battery can reach 400Wh/kg.

In terms of all solid state batteries, Funeng Technology has also made breakthroughs. In early June, the company disclosed that its research and development of all solid state batteries with an energy density of 400-500Wh/kg is progressing smoothly. The 60Ah sulfide all solid state battery is planned to be delivered in small quantities to strategic partners by the end of 2025.