Figure AI

From Humanoid Robots Wiki
Jump to: navigation, search

Figure Inc. is a robotics company focused on developing autonomous humanoid robots for industrial and potential consumer applications. Founded in 2022 in Silicon Valley, the company aims to address labor shortages by deploying general-purpose robots capable of performing complex tasks. Its current model, Figure 02, incorporates NVIDIA’s AI computing technologies, including NVIDIA H100 GPUs, enabling high-performance vision, dexterity, and conversational AI.

Figure AI
NameFigure AI
CountryUSA
WebsiteWebsite
RobotsFigure 01, Figure 02

History[edit]

Founded by Brett Adcock, Figure demonstrated rapid early progress in humanoid robotics, leveraging synthetic data from NVIDIA Isaac Sim. The company’s humanoid robots were first deployed in testing environments, such as BMW's production line, to demonstrate their real-world applications.

Technology and Partnerships[edit]

Figure’s robots utilize NVIDIA’s robotics stack, with models trained on NVIDIA DGX for large-scale AI capabilities and designed using the NVIDIA Omniverse platform. Despite having limited in-house AI expertise, Figure's collaboration with OpenAI has furthered its AI models’ conversational abilities, marking a unique blend of perception and interaction capabilities. Figure's Series B funding announcement, amounting to $675 million, supports continued research and deployment of these robotic solutions in industrial settings.

Goals and Future Directions[edit]

Figure aims to commercialize its robots for industries such as manufacturing and logistics, with a potential consumer rollout on the horizon. Their humanoid models are designed for adaptability across different environments, driven by evolving AI, perception, and mobility technologies.