机器人AI模型
Search documents
人形机器人公司Figure新融资超10亿美元,英伟达跟投
Nan Fang Du Shi Bao· 2025-09-16 15:21
Core Insights - Figure, a humanoid robotics startup, completed a Series C funding round exceeding $1 billion, achieving a post-money valuation of $39 billion, significantly up from $2.6 billion after the Series B round [1] - The Series C round was led by Parkway Venture Capital, with participation from notable investors including NVIDIA, Intel Capital, LG Technology Ventures, Qualcomm Ventures, Salesforce, T-Mobile Ventures, Brookfield Asset Management, Macquarie Capital, Align Ventures, and Tamarack Global, many of whom are existing shareholders [1] - The funding will be utilized to scale the deployment of humanoid robots in both household and commercial settings, develop next-generation GPU infrastructure, and initiate the collection of real-world datasets to enhance the robots' capabilities in complex environments [1] Company Developments - Founded in 2022, Figure raised $675 million in a Series B round in February 2024, attracting high-profile investors such as NVIDIA, Microsoft, OpenAI, and Jeff Bezos [2] - Following the termination of its partnership with OpenAI, Figure shifted to independently develop its end-to-end robotic AI models, launching the Helix model, which employs a dual-system approach for enhanced functionality [2] - The Helix model consists of a "slow system" for scene and language understanding and a "fast system" for real-time control of the robot's upper body movements [2] Product Demonstrations - Since June, Figure has showcased its robot's capabilities in logistics and household tasks, including sorting packages, loading laundry, folding clothes, and recently, loading dishes into a dishwasher [3] - The robot's ability to perform these tasks has been questioned due to the lack of public demonstrations, leading to skepticism about the effectiveness of these skills [3] - BMW is a key partner for Figure, with the robot being tested in their South Carolina plant, although there have been criticisms regarding the actual productivity and application of the robot in a production environment [3][4]