Kann Meta seine Ambitionen verwirklichen, das "Android" der Robotikbranche zu schaffen?
Recently, according to PCMag, Andrew Bosworth, the Chief Technology Officer of Meta, said for the first time at a briefing at the company's headquarters that humanoid robots have been elevated to the same strategic priority as augmented reality (AR). Meta will invest "billions of dollars" in the next few years to build a general software platform that can be licensed externally, aiming to become the "Android" of the robotics industry.
Bosworth said that Meta has no intention of mass - producing hardware. Instead, it plans to replicate Google's open - source approach in the smartphone field: any robot body that meets the technical specifications can be equipped with Meta's operating system, thereby rapidly expanding the ecosystem and gaining control over the formulation of industry standards.
Image source: PCMag
As early as February 2025, Reuters reported that Meta had established a special team within its Reality Labs to develop humanoid robots capable of performing physical tasks such as household chores.
From secretive planning to a clear statement, behind Meta's actions lies its ambition for the next - generation intelligent terminal ecosystem. However, replicating Android's success is far more difficult than expected.
01 Meta's Ambition
In Meta's view, the main problem in the current development of robots lies in the "brain" rather than the "body". Hardware technology has made significant progress in recent years, but the "brain" that enables robots to truly understand and adapt to the complex physical world and perform delicate tasks such as dexterous grasping remains the biggest challenge in the industry.
Therefore, Meta is concentrating its resources on building this "brain", constructing a "world model" to overcome software - level problems and avoid the "red - ocean" competition in hardware manufacturing.
Commercially, the platform model can effectively avoid the heavy - asset and high - risk nature of the hardware business. Manufacturing physical robots involves huge supply - chain, production costs, and inventory pressure. By providing a platform, Meta can, like offering the Android system, serve diverse hardware manufacturers at a lower cost and with higher scalability, thereby establishing an ecosystem rather than simply selling a single product.
Meta's goal is not to become a robot manufacturer but to be the "rule - maker" and the core of the underlying ecosystem in the future robotics industry.
By building a platform, it has the opportunity to set industry standards and attract many hardware partners to join its ecosystem, thus occupying the top position in the value chain of the potential huge market. This will bring a more lasting and far - reaching influence than selling hardware.
More importantly, Meta has deep accumulations in artificial intelligence, software engineering, and large - scale platform operation, but hardware manufacturing is not its traditional strength. Choosing the platform path allows it to leverage its software advantages and avoid the engineering difficulties of hardware from scratch.
Looking back at Meta's strategic transformation, in 2021, it changed its company name from Facebook to Meta and aggressively bet on the metaverse as the "next - generation Internet form". However, more than two years later, the implementation of the metaverse has fallen far short of expectations. Against this background, humanoid robots have become a key choice for Meta to find new growth points. If Meta can seize the high - ground of the robot software platform in advance, it can not only make up for the strategic gap in its metaverse business but also gain the initiative in the technological competition in the next decade.
Image source: 36Kr
02 Can It Replicate the Android Model?
The success of the Android system is a classic in the history of technology. By attracting hardware manufacturers through an open - source strategy and relying on the developer ecosystem to enrich application scenarios, it has finally formed a positive cycle from the system to hardware and then to applications. Now, Meta is trying to replicate this path to build an Android - like platform for robots.
Image source: Baidu Baike
On the positive side, Meta has the genes and resources to build an open ecosystem. The core advantage of Android is "hardware neutrality", and Meta has clearly stated that it will not be the main hardware manufacturer but only provide software blueprint licenses, which is highly consistent with Google's strategy of avoiding hardware competition and focusing on system empowerment back then.
More importantly, the current robotics industry is in a dilemma of "a wide variety of hardware but fragmented software". Industrial robot manufacturers rely on customized systems, and the software compatibility of service robots is insufficient. Developers need to repeatedly develop functions for different hardware.
This fragmented situation is very similar to the early operating - system fragmentation in the smartphone industry, providing an entry point for a general platform.
However, there are fundamental differences between robots and smartphones. The first is the problem of hardware heterogeneity. The core hardware architecture of smartphones is highly standardized, while humanoid robots involve dozens of different components such as joint motors, force - control sensors, and dexterous hands. The Android system only needs to be adapted to a limited number of chips and screen sizes, while Meta's software platform needs to be compatible with a wide range of hardware from industrial robotic arms to household service robots, and the adaptation difficulty increases exponentially.
Notably, according to Morgan Stanley data, China accounts for 63% of the global humanoid robot hardware core supply chain. If Meta wants to promote hardware manufacturers to adapt to its platform, it also needs to deal with the challenges of regional cooperation in the manufacturing end.
Image source: Tesla AI Day
Building a developer ecosystem will face long - term challenges. The success of Android is inseparable from the synergy of Google's search, app store, and other ecosystem resources. The cultivation of developers in the robotics field requires more professional toolchains. To build a competitive ecosystem from scratch, Meta not only needs to invest funds to subsidize developers but also needs to create a full - cycle toolchain covering simulation testing, safety certification, and data management.
The adaptability of the business model is also questionable. Android makes profits through app distribution and advertising, but the monetization path of robot software is not clear. The industrial scenario prefers customized solutions, and the consumer - level market still needs to cultivate users' payment habits. If Meta continues the Android's free - license model, it will be difficult to recover the "billions of dollars" in R & D investment in the short term, while a paid model may weaken its attractiveness to hardware manufacturers.
More crucially, the application scenarios of humanoid robots are not clear and highly dispersed. The core functions of smartphones are communication and entertainment, with clear and concentrated application requirements. In contrast, humanoid robots need to cover multiple scenarios such as industry, home, medical care, and agriculture, and the requirements of each scenario vary greatly, making it difficult for developers to focus and form "phenomenal applications" like in the mobile - Internet era.
That is to say, if Meta cannot wait for the market to expand, the early investment of billions of dollars may face the risk of a too - long return period. The humanoid robot industry may take 5 - 10 years to enter the explosive growth stage, which will be a huge test for Meta's capital chain and strategic patience. Coupled with the cumulative loss of more than $68 billion in its Reality Labs department in the past four years, if the robot business continues to be invested in without short - term returns, it may affect investors' confidence.
03 Formidable Competitors
However, while Meta is gearing up to make a big move, other technology players are not idle.
Google is also promoting its "Android - like system for robots" strategy and has made progress. DeepMind recently released the Gemini Robotics 1.5 series of models for robots and embodied intelligence. Gemini Robotics 1.5 can convert visual information and instructions into robot movement instructions to perform tasks, and Gemini Robotics - ER1.5 can reason about the physical world, directly call digital tools, and create detailed multi - step plans to complete tasks. At this time, Meta's robot platform has just started.
The open - source camp has a first - mover advantage. Recently, Open Mind launched the OM1 Beta system. As the world's first "AI - native" open - source platform, it not only supports platforms such as the Unitree G1 humanoid robot, Go2, Turtle Bot, and Ubtech robots but also supports the access of third - party large models such as OpenAI and Gemini. It even provides a front - end interaction interface to simplify operations. Such open - source projects quickly seize the minds of developers with the "free + collaboration" model. If Meta chooses a closed - source route, it will face a disadvantage in ecosystem expansion speed; if it follows the open - source path, it will need to bear huge R & D costs without direct benefits.
Image source: Open Mind
The vertical - integration strategies of technology giants cannot be underestimated. Tesla is adopting a visual - data - driven strategy, deeply integrating Optimus with its autonomous - driving business. The new head, Ashok Elluswamy, is in charge of both the FSD and robot projects, trying to reuse the algorithm and data advantages of autonomous driving and form a closed - loop of "algorithm - data - computing power" with the support of the Colossus supercomputer.
However, in September 2025, Ashish Kumar, a core member of Optimus' AI team, left and joined Meta, which may affect Tesla's technology - integration progress.
Image source: Cailian Press
Vertical players in the industrial and consumer sectors are also segmenting the market. Industrial giants such as Siemens and Rockwell Automation, with their in - depth understanding of high - end manufacturing scenarios, have quietly occupied the customized OS market. Their systems have become the first choice in industries such as automotive and electronics due to compliance advantages such as TÜV safety certification. The success of Android depends on the unified demand of C - end users, while the core demand for robots lies in the B - end (industrial manufacturing). Customers are more concerned about safety compliance and customized adaptation rather than a general platform. If Meta focuses on the B - end, it needs to deal with the compliance barriers of manufacturers such as Siemens and Rockwell.
04 Conclusion
Meta's elevation of humanoid robots to the same strategic priority as augmented reality (AR) is not only an accurate judgment of the trend of the integration of AI and embodied intelligence but also an inevitable choice to continue the logic of competing for "ecosystem dominance". The investment of billions of dollars demonstrates Zuckerberg's determination to replicate the Android myth, and the fragmented situation and software bottlenecks in the robotics industry also provide a theoretical possibility of success.
However, there is often a wide gap between theory and reality. Hardware heterogeneity breaks the standardized foundation of the smartphone era. Cultivating a developer ecosystem requires crossing the moat of predecessors, and the ambiguity of the business model tests the patience of long - term investment. Meta not only needs to break through technological bottlenecks but also needs to build an open, safe, and sustainable ecosystem. This race for the "next computing platform" has just begun. The dual tests of technological maturity and commercialization path will run through Meta's entire strategic implementation process, and how to find a balance among these challenges will be the key to determining whether its "Android dream for robots" can come true.
Whether Meta's "Android dream" can become a reality may take 5 - 10 years to tell. However, it is certain that Meta's entry will further accelerate the competition in the humanoid robot track. No matter who finally becomes the "Google of the robotics industry", it will promote the development of the entire industry and clear the obstacles for humanoid robots to enter thousands of households.
This article is from the WeChat official account "Shenmou Finance" (ID: chutou0325). Author: Someone concerned about humanoid robots. Republished by 36Kr with authorization.