UBTECH Robotics
Updated: November 25, 2025
Founder Zhou Jian
Country: China | Funding: $2.3B (+)
Founded: 2012
Website: https://ubtrobot.com/
UBTECH Robotics develops humanoid robots for manufacturing, service sector and households. Its proprietary ROSA platform, which combines robotics and AI technologies, enables the company to create robots with the three key characteristics of the human body (joint movement), brain (perception) and cerebellum (movement planning and control). The robot (Walker) stands 170 cm tall and has anthropometric dimensions. Thanks to a dynamic balancing algorithm similar to the human body, it can tilt 170 degrees and carry a payload of up to 15 kg. It can independently charge its batteries and schedule battery replacement or charging according to task priorities. Walker's head is equipped with a binocular computer stereo vision system with RGB illumination. It uses deep-learning-based stereo depth estimation algorithms to create highly accurate, left-aligned, dense depth maps in real time. The robot uses LLM model for interacting with people via speaker and microphone.
Founded: 2012
Website: https://ubtrobot.com/
UBTECH Robotics develops humanoid robots for manufacturing, service sector and households. Its proprietary ROSA platform, which combines robotics and AI technologies, enables the company to create robots with the three key characteristics of the human body (joint movement), brain (perception) and cerebellum (movement planning and control). The robot (Walker) stands 170 cm tall and has anthropometric dimensions. Thanks to a dynamic balancing algorithm similar to the human body, it can tilt 170 degrees and carry a payload of up to 15 kg. It can independently charge its batteries and schedule battery replacement or charging according to task priorities. Walker's head is equipped with a binocular computer stereo vision system with RGB illumination. It uses deep-learning-based stereo depth estimation algorithms to create highly accurate, left-aligned, dense depth maps in real time. The robot uses LLM model for interacting with people via speaker and microphone.
