A 4-meter-tall robot makes its debut. Are robots bidding farewell to the era of "being human-like"?

Over the past few years, the most eye-catching topic in the humanoid robot industry has been:

Can robots become more and more human-like?

Being able to walk, run, dance and perform complex actions used to be an important criterion for measuring the progress of robotics technology.

However, at the 2026 World Robot Conference (WRC), a new industry signal is emerging:

The focus of robot competition is shifting from "whether it looks like a human" to "whether it can solve real problems".

Products including the 4-meter-tall electro-hydraulic hybrid drive robot, centaur wheel-legged robot, and two-story-tall concept robot ZERO made their centralized debut, demonstrating another industrial direction: future robots may not necessarily replicate human forms, but will be redesigned based on task requirements.

This means that the robot industry is entering a new stage.

In the past, the discussion was about "whether robots have intelligence".

Now the market is starting to pay attention to:

Can robots become real productivity.

From stage performances to entering real scenarios

An obvious change released by the 2026 World Robot Conference is that robots are reducing the pure display of movement capabilities and shifting more to practical applications.

The exhibition area of this conference is nearly 50,000 square meters, with 373 enterprises bringing 3,000 innovative products to the exhibition, and 311 new products making their debut for the first time.

Among them, humanoid robots are still the focus of attention.

Unitree Technology brought humanoid robots such as G1 and H2, as well as quadruped and wheel-legged robot products, and its booth continuously attracted a large number of visitors. At the same time, more and more enterprises are beginning to demonstrate the applications of robots in scenarios such as industry, logistics, inspection, and public services.

For example:

Some robots have begun to try to enter factories to complete tasks such as handling and inspection;

Some intelligent robots enter urban public service scenarios to participate in traffic management and auxiliary patrols;

Some enterprises use remote control and data collection to let robots learn more operation capabilities in real environments.

This is significantly different from the robot display logic in the past few years.

In the past:

Robots danced, somersaulted, and completed complex actions to prove "I can move".

Now:

Robots carry goods, inspect equipment, and participate in production to prove "I am useful".

This reflects that the evaluation criteria of the entire industry are changing.

Why are robots starting to "not look like humans"?

If robots eventually need to enter complex environments such as industry, energy, mines, and rescue, then "being human-like" is not necessarily the most efficient solution.

The human body structure is suitable for the human living environment.

But for many special tasks, machines can have completely different forms.

For example:

Mine environments require stronger load-bearing capacity;

Nuclear power inspection requires radiation resistance;

Deep-sea operations require pressure resistance;

Industrial production requires stable and repetitive operation capabilities.

Therefore, more and more enterprises are beginning to explore non-traditional robot forms.

At this conference, the 4-meter-tall electro-hydraulic hybrid drive super-humanoid robot has become one of the focuses of attention.

This type of robot is not designed to simulate the life of ordinary people, but for special environments such as nuclear power operation and maintenance, deep-sea operations, mining, and emergency rescue.

Another type of robot that has attracted attention is the centaur wheel-legged robot.

It combines the dual-arm operation capability of humanoid robots and the movement capability of quadruped robots, can adapt to complex terrain, and is used in scenarios such as inspection, industry and rescue.

The logic behind this is very clear:

In the future, robots may not "replicate a human", but create a new type of tool that can complete tasks that humans cannot.

Humanoid robots still have value, but are no longer the only answer

Of course, this does not mean that the humanoid robot route will disappear.

On the contrary, humanoid robots still have natural advantages.

The reason is:

Most of the infrastructure in human society is designed according to the human body structure.

Door handles, stairs, tools, and operating spaces on production lines are all formed around human habits.

Therefore, a robot close to the human body structure can theoretically enter the existing environment more easily.

This is also why a large number of enterprises around the world are still investing in the research and development of humanoid robots.

But the problem is:

If the robot is facing a brand new task environment, then a specially designed robot may be more efficient.

For example:

Warehouse handling does not necessarily require two legs;

Mine rescue does not necessarily require a human body structure;

Industrial inspection does not necessarily need to fully simulate human movements.

In the future, the robot industry may form a multi-route competition pattern:

Humanoid robots are responsible for general scenarios;

Wheel-legged robots are responsible for complex environments;

Special robots are responsible for professional tasks.

What really determines the value is not the appearance, but the commercial efficiency.

The investment logic of the robot industry chain is changing

The development of the robot industry is also changing the focus of the capital market.

In the early stage, the market focused on robot complete machine enterprises.

But as the industry enters the application stage, value may gradually spread to core components and basic capabilities.

The robot industry chain mainly includes:

First, core executive components

Including:

· Reducer;

· Motor;

· Servo system;

· Joint module.

These components determine the power and motion capabilities of the robot.

Second, perception system

If a robot wants to enter the real environment, it needs to perceive the world like a human.

Including:

· Vision sensor;

· Force sensor;

· Tactile sensor.

Third, the artificial intelligence "brain"

The future competition of robots largely depends on:

Whether it can understand tasks;

Whether it can plan autonomously;

Whether it can continuously learn through data.

This is also why the robotics industry has increasingly emphasized embodied intelligence in recent years.

Robots are not just mechanical devices, but a new entry point for artificial intelligence to interact with the real world.

The biggest challenge: shifting from display capability to commercial value

The robot industry has huge prospects, but there are still challenges before large-scale commercialization.

The first problem is cost.

High-performance robots still require a large number of high-value hardware at present, and the commercial procurement cost is relatively high.

The second problem is reliability.

Enterprises will not buy a robot just because it "can perform", but because it can work stably for thousands of hours.

The third problem is data.

After robots enter the real environment, a large amount of real operation data is required for training.

How to quickly accumulate data and make robots adapt to different scenarios is a problem that the industry must solve.

Therefore, the real competition in the robot industry is not about who can show more complex actions in the short term.

It is about who can take the lead in establishing:

Hardware capabilities + AI model + data closed loop + commercial scenarios.

From "human-like" to "born for tasks", robots enter a new stage

In the past few years, the humanoid robot industry has gone through the stages from concept verification to technology demonstration.

Being able to walk, jump, and interact has let the market see the possibility of robot development.

But in the next few years, industrial competition will enter a more realistic stage:

Who can reduce costs?

Who can enter the production line?

Who can create real economic value?

The signal released by the 2026 World Robot Conference is:

The robot industry is bidding farewell to the stage of simply pursuing "being human-like".

The most valuable robot in the future may not be the one whose appearance is closest to humans, but the one that is safer, more efficient and more stable than humans in a specific scenario.

From imitating humans to task-driven, robots are completing an important shift.

And this shift may determine the value ranking of the robot industry in the next stage.

Risk Warning: This article analyzes industrial trends based on public information and does not constitute investment advice. The robot industry is still in a stage of rapid development, and there are uncertainties in the commercialization progress, technical routes and market competition pattern.

This article is from the WeChat Official Account "BT Finance" (ID: btcjv1), the author is BT Finance, and 36Kr is authorized to release it.