Figure Partners with Nvidia to Accelerate Robot Deployment

Share this article
Share this article
Prioritise Us on Google
Figure designs general-purpose humanoid robots. Picture: Figure
Figure's latest Series C funding supports the company’s efforts to deploy general-purpose humanoid robots into real-world environments at scale

Figure has surpassed US$1bn in committed capital during its Series C funding round, bringing its post-money valuation to an impressive US$39 billion.

This capital infusion is strategically targeted at accelerating the deployment scale of humanoid robots engineered for diverse real-world applications.

The financing round was spearheaded by Parkway Venture Capital and includes significant participation from key technology investors such as Brookfield Asset Management, Nvidia, Macquarie Capital, Intel Capital, Align Ventures, Tamarack Global, LG Technology Ventures, Salesforce, T-Mobile Ventures and Qualcomm Ventures.

According to Figure, these funds will primarily empower three technical domains: humanoid robot manufacturing and operational deployment, expansion of GPU-based infrastructure crucial for robot simulation and AI model training and extensive real-world data acquisition for enhancing embodied intelligence via its proprietary AI platform Helix.

Youtube Placeholder

Scaling manufacturing and deployment 

Central to this funding round is Figure’s commitment to increasing production capacity and commercial deployment of humanoid robots designed to operate in both residential and industrial contexts. This includes a significant ramp-up at the BotQ manufacturing facility.

BotQ acts as the high-volume production centre where Figure’s humanoid robots are engineered to replicate human dexterity and interaction capabilities within physical environments.

The company aims for these robots to perform a broad spectrum of tasks typical in sectors like logistics, warehousing, and domestic operations – particularly those involving repetitive, labour-intensive, or safety-critical activities.

As demand escalates for robotics that seamlessly merge physical automation with advanced AI reasoning, Figure is positioned to address complex operational needs requiring embodied intelligence.

Figure has progressed through two developmental generations of robots over three years, currently advancing the model named Figure 03.

Brett Adcock, CEO at Figure, says: "Figure's goal is to solve general robotics. This is a really hard problem.

Brett Adcock, CEO at Figure. Picture: Figure

"For the first time in history, the right technologies exist to make that possible: robots that can achieve human-level intelligence. Today, we're announcing over a billion dollars of funding. The team is in place, the robots are built and the path ahead is clear."

Training Helix through next-gen compute infrastructure

A significant fraction of the newly secured capital will be allocated to architecting and scaling sophisticated GPU computing infrastructure.

GPUs are essential for AI workloads due to their capability for large-scale parallel data processing, which is foundational for training complex machine learning models.

Figure’s Helix platform, its proprietary AI system, serves as the nexus of perception, reasoning, and control for their humanoid robots – encompassing what the company terms “embodied intelligence”.

Enhanced GPU capacity will exponentially improve Helix's ability to simulate environments, enabling more nuanced interpretation, planning, and task execution.

Helix operates on an integrated “vision-language-action” framework, synthesising visual object recognition, natural language comprehension and motor control to facilitate command adherence and dynamic contextual adaptation.

Figure’s ambition is to scale humanoid robots into both residential and commercial settings. Picture: Figure

Building real-world data through Brookfield partnership

In a landmark partnership, Figure has aligned with Brookfield, a global alternative asset manager overseeing more than US$1 trillion in assets and more than 100,000 residential units, to develop expansive real-world datasets vital for training Helix.

Brookfield’s vast portfolio offers diverse environments for large-scale collection of human navigation and object manipulation data, which will serve as training inputs to precondition Helix's AI for adaptability to complex, dynamic scenarios humanoid robots will encounter in deployment.

Brett explains: "This partnership marks a major milestone in our journey to build general-purpose humanoid robots. Brookfield’s scale gives us an unmatched platform to capture massive amounts of real-world, humanlike navigation and manipulation data across a variety of household environments necessary to unlock general-purpose humanoid robots."

Brookfield CEO Bruce Flatt adds: "This approach to collaboration furthers Brookfield’s position at the forefront of integrating AI to drive productivity in real assets and business."