XDOF Aims for $1.2 Billion Valuation in Upcoming Series B by Harnessing Real-World Robot Data
The AI startup is on track to enhance robotic training data collection through innovative methods.
The Full Story
XDOF, a startup specializing in real-world teleoperation data for robot training, is in late-stage talks for a Series B funding round at a $1.2 billion valuation. Founded in 2024 by UC Berkeley researchers Philipp Wu and Fred Shentu, XDOF has quickly gained traction, announcing a $70 million Series A just three months ago with backing from notable investors like Thrive Capital and Andreessen Horowitz. The startup's rapid growth, achieving annualized revenues nearing $50 million, has led venture capitalists to pursue further investment within a short time frame.
While the final terms of this deal remain under negotiation, it underscores the urgency and excitement surrounding XDOF's innovations. XDOF operates by building robust data pipelines and training systems necessary for AI labs and robotics companies. This focus is crucial, as collecting high-quality datasets for physical robots is significantly more complex than for large language models (LLMs).
Wu's previous research highlighted a critical gap in available real-world data, which inspired the creation of XDOF. The company’s flagship project, GELLO, introduced a teleoperation system that enables human operators to control robotic arms remotely, generating valuable training data. Through this approach, XDOF acts as an outsourced data-supply chain, catering to the robotic industry that lacks comprehensive datasets.
Investors liken XDOF to data-labeling giants like Scale AI, highlighting its importance in enabling robotic and AI advancements. Additionally, XDOF has partnered with UC Berkeley's AI Research lab to introduce what it claims will be the largest collection of premium robot training data ever compiled, branded as ABC. This ambitious effort combines remote robots with human data collectors equipped with sensors to record everyday activities, thereby enriching the datasets used for training robotic systems.
XDOF’s future plans include hiring and training global teams of data collectors, comprising teleoperators who manage robot tasks remotely and egocentric operators who gather movement data through wearable technology. The startup is currently collaborating with over 20 clients, including several top-tier AI labs, cementing its position in the competitive landscape of real-world data acquisition for robot training. XDOF's innovative approach aims to dismantle obstacles in training general-purpose robots by providing essential data, thus ensuring enhanced growth in the robotics sector.
As they move forward, the implications of their funding and subsequent innovations could significantly impact the future of AI and robotics technology on a global scale. Investors and enthusiasts alike will be keeping a close watch on the developments at XDOF in the coming months, as they shape the landscape of robotic capabilities and applications around the world. This Series B funding could further support XDOF in scaling its operations and enhancing its data collection techniques, allowing for a more sophisticated future in robotic technologies.
Why It Matters
XDOF's advancements could revolutionize how robots are trained using real-world data, addressing a critical gap in the industry. By enhancing data availability for robotic systems, they position themselves as integral to future technological developments.
What's Next
As XDOF finalizes the details of its Series B funding, the startup plans to expand its data collection efforts globally, potentially increasing its partnerships with more AI labs and robotics companies in 2027 and beyond. This could lead to groundbreaking developments in the field of robotics.