Hugging Face Acquires Pollen Robotics, Expands AI Portfolio with $70,000 Reachy 2 Robot

Hugging Face, a prominent AI company, has expanded its portfolio by acquiring Pollen Robotics, a French startup specializing in humanoid robotics. This acquisition introduces Reachy 2, an AI-powered humanoid robot priced at $70,000, designed for research, education, and embodied AI experiments. Reachy 2 is already in use at notable institutions such as Cornell University and Carnegie Mellon University, supporting applications in human-robot interaction, machine learning, and dexterous task execution.

Reachy 2 is distinguished by its human-like 7-degree-of-freedom arms, the proprietary Orbita joint system for expressive movement, and a mobile base equipped with Omni wheels and LiDAR. These features enable intuitive interaction and agile manipulation. Additionally, the robot supports VR teleoperation, allowing users to remotely control it and see through its “eyes” in real-time. The robot's open-source nature means that its hardware designs, software, and documentation are freely available on Hugging Face’s platform.

Ask Aime: What impact will Hugging Face's acquisition of Pollen Robotics have on the AI industry and the companies involved?

The acquisition of Pollen Robotics brings approximately 30 new employees to Hugging Face, including ex-Tesla staff scientists, signaling the company’s ambition to compete in the global robotics race. Hugging Face aims to integrate physical robotics into its open-source ecosystem, maintaining its ethos of accessibility and collaboration. The company has raised significant investment, positioning it to bring AI and robotics closer together and empower a global community of developers to innovate in both virtual and physical domains.

Hugging Face’s entry into robotics follows recent initiatives, including the hiring of Remi Cadene, a former tesla Optimus researcher, and the launch of the “LeRobot” open-source library. The company also collaborated on an affordable $100 robotic arm, further demonstrating its commitment to making robotics accessible and collaborative. The acquisition of Pollen Robotics is a significant step in Hugging Face’s vision of a future where everyone can build or use robot assistants or games, starting from open solutions instead of closed, remote-controlled hardware.

Pollen Robotics has established itself as a key player in developing humanoid robots for advanced research and real-world applications. Reachy 2, its flagship robot, is designed as a flexible, user-friendly platform for embodied AI. The robot has earned recognition, including second place in the ANA Avatar XPRIZE in 2022, and has been deployed in hundreds of units across over 20 institutions. The acquisition by Hugging Face aligns with Pollen Robotics’ core belief in the profound role robots will play in our lives, serving as the interface between AI and the physical world.

Reachy 2 is the first manufactured, commercially available, open-source humanoid robot. It is a versatile, expressive, and open robotic platform designed to explore the future of human-robot interaction, assistive robotics, and AI-driven behaviors. The robot has a “humanoid upper body” and includes stereo vision, microphones, a speaker, and lidar. Each of the robot’s arms has seven degrees of freedom, allowing it to move in seven independent directions, mimicking the flexibility of a human arm. The robot runs on Robot Operating System 2 (ROS 2) and Hugging Face’s LeRobotHF framework and is accessible via a Python-based software interface.

Hugging Face’s co-founder, Clément Delangue, clarified that Reachy 2 is not intended for consumer use but is specifically designed for AI labs. The company’s broader ambitions for Reachy 2 and the future of open-source robotics include a vision where everyone from the community can build and control their own robot companions instead of relying on closed, expensive black boxes. While Reachy 2 is currently aimed at research institutions, Hugging Face’s investment in humanoid robotics suggests that the machines could play a larger role in the evolution of open-source AI.