Collection: DYNAMIXEL-Q
DYNAMIXEL-Q
Powers AI Sapiens K1
DYNAMIXEL-Q is ROBOTIS’ QDD actuator platform for dynamic humanoids, contact-rich motion, torque-aware control, and next-generation Physical AI.
The new AI Sapiens K1 demonstrates what DYNAMIXEL-Q is built to enable: a real 23-DOF humanoid platform for locomotion, whole-body control, reinforcement learning, imitation learning, and sim-to-real development.
DYNAMIXEL-Q and AI Sapiens K1 information is provided as an early product and platform preview. Specifications, software behavior, open-source resources, documentation, model names, and availability may change without prior notice.
Whole-body humanoid motion platform.
Compact physical AI research platform.
Humanoid form factor for real-world development.
DYNAMIXEL-Q actuator system voltage.
Built for Real Robot Dynamics
Physical AI requires hardware that can move through contact, respond to disturbances, and execute learned behaviors on real machines. DYNAMIXEL-Q combines QDD actuator design, low gear reduction, integrated electronics, high-speed communication, and torque-aware control for robots that need more than simple position movement.
- Quasi-Direct Drive architecture for dynamic humanoid motion
- Low-inertia PMSM actuator design for responsive control
- Low backlash and torque-aware behavior for contact-rich robotics
- Integrated actuator platform for real-time whole-body control
- Used across AI Sapiens K1 for locomotion, posture, and upper-body motion
A Humanoid Platform for Physical AI
AI Sapiens K1 brings DYNAMIXEL-Q actuators, robot software, simulation assets, training workflows, and deployment guides into one humanoid platform. It is designed for researchers, educators, and developers working on locomotion, motion generation, reinforcement learning, imitation learning, and real-world deployment.
Arms, legs, and waist for full humanoid movement.
Develop against real hardware, not simulation only.
Hardware, software, simulation, and learning resources.
One Platform for Learning, Control, and Deployment
AI Sapiens K1 connects simulation, training, control, and real-world execution so users can develop policies and validate them on DYNAMIXEL-Q powered hardware.
Develop motion and policies in simulation using robot assets and physics-based training workflows.
Use reinforcement learning and imitation learning workflows for locomotion and repeatable robot behavior.
Transfer commands through ROS 2 control to DYNAMIXEL-Q actuators on the real humanoid.
Test against real dynamics, contact, balance, and actuator feedback for practical Physical AI iteration.
The Q-Series Hardware Behind the Humanoid
AI Sapiens K1 uses two DYNAMIXEL-Q actuator sizes: a compact QC060 model for smaller joints and a higher-capacity QC080 model for larger load paths.
Compact QDD actuator for joints where lower mass and smaller packaging are important.
| Motor | Low-inertia inner rotor PMSM |
| Rated Voltage | 48 V |
| Ratio | 20:1 |
| Max Torque | 30.0 Nm |
| Max Speed | 200.0 rpm |
| Size / Weight | Ø58 × 60 mm / 435 g |
Larger QDD actuator for K1 joints that require higher torque capacity and larger load paths.
| Motor | Low-inertia inner rotor PMSM |
| Rated Voltage | 48 V |
| Ratio | 20:1 |
| Max Torque | 72.6 Nm |
| Max Speed | 110.0 rpm |
| Size / Weight | Ø78 × 65 mm / 845 g |
AI Sapiens K1 Hardware Overview
| Category | Specification | Why It Matters |
|---|---|---|
| Model | AI Sapiens K1 | Humanoid platform for Physical AI development. |
| Height | 1366 mm | Compact humanoid size for lab and research environments. |
| Weight | 35 kg | Full-body platform with real actuator and contact dynamics. |
| Degrees of Freedom | 23 total | Supports locomotion, posture, balance, upper-body motion, and waist coordination. |
| Joint Layout | 5 DOF per arm, 6 DOF per leg, 1 waist DOF | Whole-body structure for humanoid motion research. |
| Actuators | DYNAMIXEL-Q QC060 ×12, QC080 ×11 | QDD actuator system for dynamic humanoid control. |
DYNAMIXEL-Q in a Real Humanoid Platform
AI Sapiens shows the type of motion, control, learning, and real-world robot behavior that DYNAMIXEL-Q is being developed to support.
Train Locomotion, Then Validate on Hardware
AI Sapiens supports reinforcement learning workflows for locomotion development, helping users iterate in simulation and reduce the gap between policy training and real humanoid deployment.
From Human Demonstration to Robot Behavior
Demonstration data can be collected, structured for training, and transferred back to the robot so learned behaviors can be tested as natural, repeatable motion on DYNAMIXEL-Q powered hardware.
Get Ready for DYNAMIXEL-Q and AI Sapiens K1
DYNAMIXEL-Q is being developed for the next generation of Physical AI robots: humanoids, legged systems, dynamic manipulators, and real-world learning platforms. AI Sapiens K1 brings that actuator platform into a complete humanoid development system.