Apple, Meta, Google... Who will create the "Android" system of the humanoid robot era?

The global humanoid robot market is on the brink of explosion. Who will create the "Android" system of the humanoid robot era?

According to news from the Chasing Wind Trading Desk, Morgan Stanley stated in its latest research report that large tech companies are competing for ecological dominance in the humanoid robot field. In the past month, giants such as Apple, Meta, Google, and Huawei have successively released open-source robot models and tools, aiming to attract developers to their respective robot ecosystems.

At the same time, the Chinese government continues to strengthen its support for the humanoid robot industry, with the scale of investment funds reaching 187 billion yuan. Morgan Stanley analyzes that manufacturing capability and physical intelligence are closely intertwined, and data acquisition will determine the ultimate winners of AI-driven robotic technology. China's advantages in manufacturing foundation and policy support may give it a leading position in this "robot race."

According to the Morgan Stanley report, by 2050, the global humanoid robot market's annual revenue could reach $5 trillion, with cumulative adoption reaching 1 billion units. This scale is approximately twice the total revenue of the world's top 20 automotive manufacturers in 2024.

As of the market close on June 24, the "Humanoid Robot 100" index has risen 14.4% this year, outperforming the S&P 500 index by about 11 percentage points. Among them, Shuanglin Co., Ltd. saw a staggering increase of 148%, ranking first, while Jinli Permanent Magnet rose by 132%. Chinese companies are receiving significant attention, reflecting China's strong position in the humanoid robot industry chain.

Tech giants' open-source competition heats up, future "Android system" winner yet to be determined

In the past month, tech giants have significantly accelerated their open-source layout in the robotics field. These companies hope to extend AI tools into the physical world through open-source strategies while binding developers more deeply to their respective ecosystems.

• Meta released the V-JEPA 2 world model, an AI world model with 1.2 billion parameters that can "understand, predict, and plan the physical world."
• Google DeepMind launched Gemini Robotics On-Device, allowing the company's Gemini robot models to run locally on robots to reduce latency and achieve integrity.
• Apple released the EgoDex dexterous manipulation dataset, claiming it to be the largest and most flexible dexterous manipulation dataset released to date.
• Huawei released the CloudRobo embodied intelligence platform and proposed the Robot-to-Cloud (R2C) protocol, hoping to enable more robots to develop standardized and secure standards.

Morgan Stanley pointed out:

As companies diversify their AI investments into the physical world, including hardware, the pace of development will only accelerate further. Notably, many tech companies have only recently begun large-scale recruitment of robotics talent.

Unprecedented support from Chinese policies, automotive companies have natural advantages in transitioning to humanoid robots

Morgan Stanley stated that similar to the support for electric vehicles, batteries, and drones, China's support for the humanoid robot industry has reached unprecedented heights.

• Following the Beijing Humanoid Robot Half Marathon, the world's first humanoid robot boxing match was held in Hangzhou and broadcast nationwide on CCTV.
• Government-level investments are also substantial. Various levels of government have established industrial funds with a total scale of approximately 187 billion RMB to support the development of the humanoid robot value chain. This includes a 10 billion RMB robot industry development investment fund in Beijing and a 10 billion RMB AI and robot industry fund in Shenzhen.
• Wuhan City recently announced the establishment of a 1 billion RMB humanoid robot industry fund and will provide bilateral subsidies for humanoid robot deployment: application parties can receive a maximum subsidy of 1 million RMB (30% of project investment), and product providers can receive a maximum subsidy of 1 million RMB (20% of R&D investment).
• In May, financing activities for Chinese humanoid robot companies reached a record 25 transactions, with Yushu Technology completing a Series C financing round, achieving a valuation of 1.7 billion USD, with investors including Geely, Ant Group, Tencent, Alibaba, and others.

The report pointed out that China's advantages in manufacturing and intelligence should not be underestimated. Manufacturing and physical intelligence are closely intertwined, and the ultimate barrier to AI-enabled robotic technology is data acquisition, which is directly related to manufacturing capabilities. The outcome will be determined by the ability to build and deploy data physical nodes (robots) to collect data and drive the AI improvement flywheel.

Morgan Stanley's China automotive team believes that Chinese automotive companies are preparing to transition to humanoid robots, identifying humanoid robots as one of the three major challenges for traditional automotive businesses, along with strengthening automotive growth and developing smart vehicles.

Acceleration of Commercial Deployment, Expected to Reach $5 Trillion by 2050

Morgan Stanley cited a series of commercial deployment cases:

Foxconn and NVIDIA are discussing the deployment of humanoid robots in a newly built AI server manufacturing plant in Houston, with an agreement expected to be finalized in the coming months. The robots will be put into use when the factory begins producing NVIDIA GB3000 servers in the first quarter of 2026.

Amazon is developing the necessary systems to enable humanoid robots to operate as delivery workers and will begin actual testing in the near future. The tests will include various robots such as Unitree G1, which will be delivered to distribution locations via Rivian electric vans.

Hyundai Motor is expected to deploy Atlas humanoid robots at its U.S. plant, while Tesla continues to anticipate producing "thousands" of Optimus humanoid robots for deployment in its own factories by the end of the year.

Morgan Stanley stated that these deployment plans indicate that humanoid robots are rapidly moving from the proof-of-concept stage to actual commercial applications. Particularly in manufacturing and logistics, humanoid robots have achieved significant efficiency and cost reductions According to Morgan Stanley's global humanoid robot TAM model, the annual revenue of the global humanoid robot market is expected to reach nearly $5 trillion by 2050, with approximately 1 billion humanoid robots deployed cumulatively. This scale is about twice the total revenue of the top 20 global automobile manufacturers, which is projected to be $2.5 trillion in 2024.

By revenue classification: It is predicted that about 23.7 million humanoid robots will be adopted by 2036; by 2040, this number will reach 134.4 million; and by 2044, there will be 428 million humanoid robots in operation. The average selling price in high-income countries is expected to decrease from $200,000 in 2024 to $50,000 by 2040.

Overall, the humanoid robot market is at a historic turning point, driven by ecological competition in the technology industry, government policy support, the implementation of commercial applications, and recognition from the capital market, all contributing to the rapid development of this trillion-dollar emerging industry