Kathmandu. One of the biggest “wars” in the world of technology today is over who builds human-sized robots. While Tesla is focused on the hardware of its “Optimus” robot and Apple’s robotics plan is being discussed, Meta Platforms has come up with an entirely different strategy. Meta has announced plans to develop an AI software platform that will run robots rather than build them themselves.
Why is software the main challenge? Meta{
in the path to Android
According to Andrew Bosworth, CTO of Meta, the biggest obstacle in the path of robotics is software. Meta understands this and aims to build a strong AI software platform for robotics, similar to Google’s development of the Android ecosystem in the mobile sector.
Strategy: Meta has developed a prototype robot called ‘Metobot’. However, its main source of income will be in this AI software license. Which can be used by any robotics manufacturer.
Benefits: This “platform-first” strategy protects Meta from the risk of producing expensive hardware and not selling it.
This approach is similar to the lessons learned from Meta’s massive investment in virtual reality. Where software and AI cost more than hardware sales.
‘world model’ in the brain: human-like robots
At the core of Meta’s robotics platform is a “world model”. Which works as the brain of a robot.
Functionality: This AI model can interpret complex physical environments and mimic real physics. For example, it can handle objects carefully or dynamically manipulate objects like humans in a cluttered environment such as a home.
{{TAG_OPEN_span_31} Advanced Sensor Integration: Meta combines LiDAR, cameras, and ultrasonic systems with machine learning and reinforcement learning algorithms. This enables the robot to map its surroundings instantly and make quick decisions. Which is much more advanced than the traditional robotic system.
Competition and investment race
}
Meta has invested heavily in achieving this ambitious goal. The company plans to invest $460-465 billion in AI infrastructure by 2025. A portion of this investment has been earmarked for the robotics program.
Team Expansion: Meta is actively recruiting experts in AI, robotics, hardware, and software. Its infrastructure consists of thousands of GPUs and AI accelerators. This will help train advanced AI models.
Beating the competition, Meta has partnered with Germany’s Circus SE to deploy its first CA1 autonomous robot at its Munich office in September 2025.TAG_CLOSE_span_28 TAG_OPEN_span_28
This overarching strategy from Meta is part of the company’s larger vision to bring AI intelligence to the physical world. It aims to accelerate innovation — from cleaning homes to complex industrial automation — that rely on robotics. Like Qualcomm, Meta is not in chips or hardware, but in the race to dominate the software ecosystem.
