Synergistic development is the key to breaking through in the robotics industry.

As the second half of the year begins, the narrative logic for robots is increasingly focusing on practical applications, with major global robot manufacturers accelerating the mass production and implementation of their robots.

However, at such a critical juncture, Ubtech's recent mass production and delivery have sparked considerable controversy, with Figure founder Brett Adcock even stepping in to voice his doubts directly.

Brett Adcock Analyzes Frame by Frame, Raises Public Doubts

The story begins with Ubtech releasing a video last week showing the mass production and delivery of its industrial humanoid robot, Walker S2. The video shows large-scale Walker S2 units arranged in formation; these robots can collectively open their rear battery modules to demonstrate their battery-swapping capability, and they can also march in neat rows into shipping containers.

Upon seeing the video, Figure's founder immediately began a frame-by-frame analysis and concluded on X that "this is likely computer-generated imagery." His reasoning was as follows: the robots in the front row are real, but their reflections differ from those of the robots behind them. The rear robots reflect numerous ceiling lights, whereas the front row does not; this discrepancy serves as evidence of CGI.

Figure, the world's most valuable robotics company, announced on September 16 that it had completed a Series C funding round exceeding $1 billion, bringing its valuation to approximately two hundred seventy billion yuan. Consequently, its founder's public questioning of UBTECH sparked discussions among many overseas experts. Well-known robotics blogger TheHumanoidHub commented: "The video would have been better if it included scenes of humans walking and demonstrating."

In response to skepticism from Figure's founder and others, UBTECH released two videos to clarify that the mass production of the Walker S2 was not computer-generated. In the first self-proving video, staff members walked through the formation of robots while the camera continuously scanned the factory floor filled with units, accompanied by the caption, "Even I started to doubt whether I had entered a sci-fi movie."

In the second video, they directly posted an original-audio, real-time, single-take clip of the assembly process. A drone flew through the robot formation, showing the Walker S2s neatly arranged in the factory and marching forward in unison.

As of now, Brett Adcock has not issued any further response.

In fact, prior to this incident, UBTECH announced on November 10 that it had secured a 159 million yuan order for Walker S2 units, with deliveries expected to be completed by November. The company stated that its annual orders for humanoid robots exceeded 800 million yuan.

Undercurrents in the Competition for Mass Production and Implementation of Robots

Since the robot industry achieved significant breakthroughs in motion capabilities late last year, expectations for mass production and deployment have risen steadily. This has led to a prevailing sentiment in many video comment sections: 'Let's wait until robots are truly deployed; what good is performance if it doesn't land?' This reflects users' strong anticipation for the mass production and real-world application of robots.

So the main theme of the industry in the second half of this year is to accelerate the promotion of mass production and implementation. Take Figure, for example; last month they just launched Figure 03, with a plan to mass-produce 100,000 units over the next four years.

Tesla's plan is to first build an Optimus production line with an annual capacity of 1 million units in Fremont by 2026. After mass production, the cost per robot is expected to be controlled within twenty thousand US dollars, followed by the construction of a production line with an annual capacity of ten million units.

Xiaopeng, which recently launched its all-new IRON, has also planned to mass-produce advanced humanoid robots on a large scale next year.

For the robotics industry, next year is considered a critical window for the mass production and deployment of humanoid robots. Currently, major manufacturers are accelerating efforts to enhance robot intelligence to ensure that mass-produced robots possess certain advanced intelligent capabilities; otherwise, it will be difficult to find suitable users willing to pay after they hit the market.

In fact, in the early stages of industrial development, if one enterprise takes the lead in technology or mass production, this should instead inspire enthusiasm across the entire industry.

Robots represent the blueprint for the future, and there is no issue with that. However, we are still in the early stages of industry development. Issues such as data, models, and production costs remain significant obstacles preventing robots from entering households. This requires global collaboration within the robotics industry to overcome technical challenges together.

Therefore, the past year has seen numerous open-source initiatives in datasets, models, and other technologies, all aimed at accelerating the development of the robotics industry.

As a globally co-built industry, collaboration and symbiosis, rather than vicious competition, are key to expanding the market and creating a virtuous cycle. Perhaps at some point in the future, we will realize that the current fierce struggles hold little significance.