Back to Tokyo Robotics
Torobo

Torobo

Torobo

Last updated: September 27, 2026
Locomotion Type
Wheeled
Stage
Demo
Height
161.5 cm
Degrees of Freedom
24 DOF
Weight
120 kg
Max Speed
5 km/h
Dual-Arm Payload
14 kg
Battery Life
3 h
Release Year
2024
Computing
Onboard PC as standard; optional second image-processing PC with NVIDIA RTX 5070
Sensors
Optional sensor head: wide-angle stereo camera x1, fisheye camera x1, ToF depth camera x1, stereo microphone x1, speaker x1; torque sensors at all arm and waist joints

Torobo is the signature humanoid robot of Tokyo Robotics, designed for manufacturing, logistics and service applications and currently labelled 'Under Development' by the company. It is built at close to human scale, standing 1615 mm with a 590 mm cart width and weighing approximately 120 kg, and its worst-case single-arm payload is 7 kg. Its joint layout combines 7-axis dual arms, a 3-axis waist, a 3-axis neck and a 4-axis omni-directional mobile base, a total of 24 axes, which gives it a range of motion comparable to a person in living and working spaces. Torque sensors at every arm and waist joint drive impedance control, enabling safe contact stops and force-controlled execution of assembly, cooking and physical human interaction. The head and end effector are options: the head can be a plain version or a sensor head with a ToF depth camera, fisheye camera, wide-angle stereo camera and microphones plus a speaker, while the end effector can be an in-house gripper with current-limited gripping force and back-drivable operation, or a multi-fingered hand. Computation is a standard onboard PC with an optional second image-processing PC for machine-learning training and inference, and the robot runs up to about three hours on a built-in battery with continuous operation while charging. The whole stack is ROS-based, with Cartesian impedance control, ZMP-based fall prevention, movement state machines and interference-detection safety stops, plus RViz and MoveIt! support, and models are published for MuJoCo and NVIDIA Isaac Sim so behaviour can be validated in simulation first.

Application Scenarios

Manufacturing; logistics; service; machine-learning research on robots; tasks requiring active contact with objects or the environment, such as assembly and cooking

Subscribe to Weekly Report

Get weekly updates on humanoid robot funding, new companies, and product launches.