Democratizing Humanoids: Modular, Open, and Accessible for Every Developer with David Zhang
David Zhang, founder of Highto Robotics, presents desktop humanoid robots and modular components at the Humanoids Summit with Mini Pi and P Plus.
In this talk
- 2:09
Robots built for everyday development
Zhang says his company aims to bring personable robots into daily life, rather than confining them to factories. He compares the current moment in robotics to the era when smaller, more accessible computers opened up development to more people. He sees compact robots as platforms for developers and researchers while robot intelligence is still limited.
- 7:17
A modular platform with personality
Zhang describes a future where robots offer both development tools and engaging character features. Their modular design lets users change a robot’s shape and configuration, from bipedal and humanoid forms to quadrupeds, six-legged machines, and robotic arms. He gives a character created for China’s National Games, the Great Bay Chicken, as an example.
- 9:16
Components for building and controlling robots
Highto Robotics offers more than complete robot bodies: its system includes a control box, battery, gripper, actuators, and an open-source arm. Zhang says the newer UniBox has seven CAN bus channels, each able to control 10 actuators at 1,000 Hz or 20 at 500 Hz. The company began making compact actuators after it could not find suitable ones for its small robots.
- 14:17
Accessibility, pricing, and partners
Zhang calls accessibility central to realizing robotics’ wider value. He says P Plus was available for preorder at about $5,500, with a $5,000 summit price, and that shipping was expected to begin in January 2026. He also said the company planned to begin a seed funding round around that time and was looking for senior partners.
- 17:56
Keeping learners engaged
Asked how to prevent students from losing interest, Zhang points to soccer competitions and the experience of watching and participating, alongside the hardware. He says students can develop their own algorithms using reinforcement learning and imitation learning. Distinct approaches and robot personalities can make competitions more engaging over time.
Briefs summarise talks and interviews; timestamps open the video at that moment. Claims in a talk are the speaker's own.