China leads the body, NVIDIA wants to be the brain, Tesla "integrates" — Morgan Stanley elaborates on the top 100 humanoid robot companies!

Wallstreetcn
2025.02.07 03:42
portai
I'm PortAI, I can summarize articles.

Morgan Stanley systematically sorted out 100 core listed companies in the global humanoid robot industry chain for the first time, focusing on three core links: "brain," "body," and "integrator." It pointed out that China occupies 63% of the humanoid robot supply chain, holding a dominant position, especially with significant advantages in the "body" link, while technology giants like Tesla and NVIDIA have become key players in various links

Author: Li Xiaoyin

Source: Hard AI

At Tesla's latest earnings conference, Musk stated that the Optimus humanoid robot has enormous commercial potential and could generate over $10 trillion in revenue in the future, becoming a major source of income for the company.

This statement highlights the importance of humanoid robots as a significant potential market, as the global humanoid robot industry may be 迎来 its "ChatGPT moment."

Morgan Stanley analysts Adam Jonas, William J Tackett, and others stated in their latest research report that humanoid robots will become one of the biggest themes in technology investment over the next decade, estimating that the total addressable market (TAM) for humanoid robots could reach $60 trillion, comparable to the scale of the global economy.

In the report titled "Humanoid 100," Morgan Stanley systematically reviewed 100 core listed companies in the global humanoid robot industry chain for the first time, pointing out that China dominates the supply chain in this industry, while technology giants like Tesla and NVIDIA become key players in various segments.

Industry Chain Breakdown: Brain, Body, Integrators

Morgan Stanley divides the humanoid robot industry chain into three core segments: "Brain," "Body," and "Integrators."

  • "Brain" (22 companies): Covers AI chips, software, and semiconductors, serving as the intelligent core of the robot.

The report states that the "brain" of humanoid robots mainly consists of semiconductors and software, with the most critical components being the foundational generative AI models for autonomy, as well as simulation models and digital twin technology for training.

The report lists 22 "brain" companies, including technology giants like NVIDIA, Microsoft, Google, and Meta, as well as chip design companies like Arm, Synopsys, and Cadence.

Among them, key players include NVIDIA, Microsoft, and Google, which will dominate generative AI models and computing power chips, while Baidu and Horizon and other Chinese companies will accelerate their catch-up in the visual chip field.

Morgan Stanley points out that AI algorithms can enhance the operating systems of humanoid robots by improving scalability, accuracy, and stability. In the long term, companies with leading operating systems may dominate the direction of technological advancements in humanoid robots and the pace of large-scale production.

  • "Body" (64 companies): Includes hardware components such as sensors, batteries, motors, and reducers.

The report states that the "body" of humanoid robots mainly consists of sensors, actuators, a network of wires and connectors, and lithium-ion batteries, with the exterior typically made of a mixture of aluminum alloy and plastic to minimize weight.

Among them, the key components of humanoid robots and the main companies involved include:

Actuators: Composed of screws, reducers, motors, sensors, bearings, and encoders. Tesla's Optimus Gen2 uses 28 actuators to drive 50 degrees of freedom, involving companies such as Harmonic Drive (Japan) and Hengli Hydraulic (China).

Sensors: Including visual sensors, force sensors, inertial sensors, temperature sensors, etc. Six-axis force sensors are the most complex type of force sensor. It is expected that Sony and Hesai Technology will dominate the production of visual and LiDAR sensors.

Batteries: Typically cylindrical lithium-ion batteries, installed in the center of the robot's torso. It is expected that CATL and LG Energy will dominate battery supply.

Analog semiconductors: Used in conjunction with sensors to help regulate temperature, speed, position, and power distribution.

  • "Integrators" (22 companies): Companies developing complete humanoid robots, mainly automotive and tech giants.

Morgan Stanley states that "integrator" companies mainly include four categories: automotive manufacturers, consumer electronics companies, e-commerce and internet companies, and traditional robot manufacturers:

Automotive companies: Including Tesla, Hyundai/Boston Dynamics, Toyota, etc., leveraging advantages from the electric vehicle supply chain.

Consumer electronics companies: Including Apple, Xiaomi, Sony, etc., accelerating the layout of home scenarios.

E-commerce and internet companies: Including Alibaba, Amazon, Tencent, etc.

Traditional robot companies: Including ABB, Midea (KUKA), etc.

Chinese Supply Chain: Cost Advantages and Government Support are Key

The report shows that currently, 73% of the companies confirmed to be involved in humanoid robot development and 77% of integrators are located in Asia (of which 56% and 45% are from China, respectively). Among them, China holds a 63% share in the humanoid robot supply chain, especially with significant advantages in the "body" segment:

Actuator components: Hengli Hydraulic (601100.CN), Shuanghuan Transmission (002472.CN), and other companies have entered Tesla's supply chain; Rare Earth Magnetic Materials: China controls 90% of global rare earth processing, with Ningbo Yunsheng (600366.CN) and Northern Rare Earth (600111.CN) providing core materials for motors.

Morgan Stanley expects that future startups will benefit from mature supply chains, local application opportunities, and strong government support. It is estimated that by 2050, China's humanoid robot market size will reach 6 trillion yuan, with a total of 59 million humanoid robots.

In contrast, European and American companies are more focused on the "brain" segment. The report states that NVIDIA is building an ecosystem with AI chips and the simulation platform Omniverse, while Tesla is transitioning to robotics through its FSD (Full Self-Driving) algorithm, forming a soft and hard integration advantage