Following its separation from OpenAI and coinciding with its push for Series C funding, Figure AI has unveiled its second-generation robot Figure 02.
At the heart of Figure AI’s announcement is Helix, a trailblazing Vision-Language-Action (VLA) model that powers the Figure 02 robot and is engineered for generalist humanoid control.
Despite the possible hype in generalist robot development, we identify four major factors that justify this leap – breakthrough technology IP, market expansion, adaptive generalization and commercial scalability.
Figure AI, a rising robotic unicorn, is trying to rewrite the playbook on humanoid robotics. In a bold strategic pivot that follows its separation from OpenAI and coincides with its push for Series C funding, the company unveiled its second-generation robot Figure 02 on February 20, followed by a series of demo videos in March.
Key technical breakthroughs
At the heart of Figure AI’s announcement is Helix, a trailblazing Vision-Language-Action (VLA) model that powers the Figure 02 robot and is engineered for generalist humanoid control:
Four Key Technical Breakthroughs With Helix
Source: Counterpoint, Figure AI
We pulled key screenshots from the demo video, highlighting two robots — both running on a single instance of Helix — collaborating in real time to organize groceries. The catchS? They have never seen these items before. This is not scripted automation; it is adaptive intelligence in action.
Demo Video Showcasing the Technical Breakthroughs With Helix
Source: Screenshots from the Figure AI Official Demo Video
Valuation surge: 15.2x growth
Along with Figure 02/Helix, Figure AI is actively pursuing a new round of funding that could skyrocket its valuation to $39.5 billion, a remarkable 15.2x leap from last year.
Figure AI’s Funding History With Valuation and Key Investors
Source: Company Announcements, Counterpoint Research
Despite the possible hype in generalist robot development, we identify four major factors that justify this leap:
Breakthrough technology IP: Helix is a cutting-edge AI solution that has advanced end-to-end reasoning capabilities. By moving away from OpenAI’s AI (embedded in Figure 01) and building Helix in-house, Figure AI has eliminated third-party dependence on a separate AI provider, gaining full control over its core AI technology. IP ownership makes Helix a proprietary asset that contributes to Figure AI’s whopping valuation.
Market expansion: Extending focus from factory-centric applications to home robotics taps into a vast, underexplored market. Figure AI’s pivot to home robotics represents a decisive competitive advantage in an industry increasingly crowded with high-profile entrants, with Elon Musk’s Optimus and China’s Unitree making strides in factory or industrial robotics settings.
Adaptive generalization: The inherent challenges of home environments, where objects vary in size, shape and texture, underscore the need for advanced generalization capability. Helix’s ability to learn new objects on the fly through voice commands, coupled with its multi-robot collaboration, offers an edge in this scenario, sparking plenty of imagination for future home applications among investors.
Commercial scalability: Figure AI claims Helix is the first Vision-Language-Action (VLA) model engineered to run exclusively on embedded, low-power GPUs — completely sidestepping the need for cloud infrastructure and achieving commercial readiness. More than just reducing costs, this efficiency is a game changer for mass production. That is also why Figure AI is setting an ambitious target of shipping 100,000 humanoid robots within the next four years.
To learn more about Counterpoint Research's AI data and insights, explore our AI 360 service here: https://www.counterpointresearch.com/coverage/coverage-ai.
Related Research
Dec 19, 2024
Jul 27, 2024