Image/s credit: Rainbow Robotics.

Rainbow Robotics RB-Y1 Wheeled Humanoid

company logo for Rainbow Robotics Rainbow Robotics

Country: South Korea

Year: 2024

Summary

RB-Y1 is a wheeled humanoid research platform that combines two 7-DoF arms, an articulated height-adjustable body and a fast mobile base for AI, bimanual manipulation, and more.

Status: Research, Commercially available

Operation: Semi-autonomous, Teleoperation

Robot Type / Domain: Multi/General-Purpose, Research, Industrial, Service

Tasks: Bimanual manipulation, assembly research, parts handling, pick-and-place

More Videos

Play video background image - video 2
Amazing Drawing and Painting by RB-Y1
Play video background image - video 2
Imitation Learning and Teleoperation with RB-Y1
Play video background image - video 2
15 Years of Amazing Robotics with Rainbow Robotics

Learn More

A Humanoid Designed to Move on Wheels

RB-Y1 is best understood as a humanoid-style mobile manipulator rather than a conventional legged humanoid. It has two articulated arms and a human-like working arrangement, but its lower body consists of a single articulated support mechanism mounted on a wheeled platform. This design allows it to move efficiently across level floors without having to solve the difficult balance, impact and energy problems associated with bipedal walking.

Its 6-DoF central mechanism performs some of the role that a person’s waist, hips, knees and ankles would perform. It can change the height and angle of the upper body, extend the robot’s working range and shift its centre of gravity during movement. Rainbow Robotics reports more than 500 mm of vertical adjustment, allowing the arms to reach objects at different heights without relying only on shoulder and elbow motion. The standard mobile platform reaches a published speed of 1.5 m/s, while an optional Mecanum-wheel configuration adds multidirectional movement for confined laboratories and industrial spaces.

Built for AI, Teleoperation and Robot Learning

RB-Y1 is supplied as a development platform rather than a finished autonomous worker with a fixed set of behaviours. Researchers can access its arms, articulated body and mobile base through official Python and C++ APIs. Rainbow Robotics also provides robot models, simulation tools and ROS 2 resources, making it possible to develop and test controllers before transferring them to the physical robot.

One of its most notable options is the 14-DoF master arm. An operator moves this separate teaching device, and the corresponding motion can be transferred to RB-Y1. This supports real-time teleoperation while also helping researchers collect human demonstration data for imitation learning. Rainbow Robotics has additionally demonstrated touchpad, joystick and virtual-reality control methods.

The software architecture separates safety-critical robot control from the user’s application computer. High-level software can therefore handle perception, task planning or AI inference while the internal robot controller manages lower-level motion. This modular arrangement makes RB-Y1 suitable for developing learned manipulation, mobile pick-and-place, vision-based control and coordinated whole-body behaviours.

Research and Potential Workplace Applications

RB-Y1 can be used to investigate tasks that are difficult for fixed single-arm robots. Its mobile base allows it to move between work areas, while the two arms can hold an object with one gripper and manipulate it with the other. Demonstrations have included cup stacking, pouring liquids and coordinated manipulation, but these should be treated as demonstrated capabilities rather than guaranteed ready-to-deploy applications.

Potential areas include two-handed assembly, parts handling, laboratory automation, machine tending, inspection, logistics research and mobile service tasks. The combination of accurate arms and an adjustable body also allows researchers to test manipulation across shelves, benches and equipment positioned at different heights.

However, RB-Y1 still requires application-specific software, sensors, end effectors and safety engineering. A vision camera is not included in the standard specification, autonomous navigation functionality depends on the selected hardware and software configuration, and the manufacturer does not publish a collaborative safety certification for unrestricted operation beside people. It is therefore more accurate to present RB-Y1 as a highly configurable research and integration platform than as a general-purpose autonomous worker delivered ready for any workplace.

DO NOT COPY

Do Not Copy

Brief History

Rainbow Robotics first presented RB-Y1 publicly at the Smart Factory + Automation World exhibition in South Korea in March 2024, before releasing detailed specifications in April. Pre-orders opened on 8 May 2024, and the robot made its overseas debut at ICRA 2024 in Yokohama, where Rainbow demonstrated real-time teleoperation using its teaching arm and simulation system. The company planned the first deliveries for October 2024. At ICRA 2025, Rainbow expanded the platform with a Mecanum-wheel system for omnidirectional movement and an integrated SDK supporting modules such as IMUs, grippers and LiDAR. Rainbow also reported that RB-Y1 units were being used by universities including MIT, UC Berkeley, the University of Washington and Georgia Tech, as well as industry users. The robot draws on Rainbow’s earlier humanoid development experience while incorporating components and engineering approaches from its collaborative-robot and autonomous-mobile-robot work.

Top 5+ Qualities (by addoobot)

What we find to be the top qualities of Rainbow Robotics RB-Y1 Mobile Humanoid:

  • Human-Like Bimanual Dexterity: Two 7-DoF arms allow independent or coordinated manipulation with up to 3 kg carried by each arm.
  • Articulated Whole-Body Movement: The 6-DoF central torso allows the robot to change its posture, working height and centre of gravity.
  • Two Mobility Options: Buyers can use the lighter differential-drive Model A or the omnidirectional Mecanum-wheel Model M.
  • High Arm Precision: Published arm repeatability of less than ±0.05 mm supports precise manipulation and research tasks.
  • Built for Robot-Learning Data: The optional Leader Arm and VR-based interfaces help researchers collect bimanual demonstrations for imitation learning.
  • Developer-Friendly Software: Official Python, C++, ROS 2 and simulation resources give researchers access to both high-level commands and real-time control.
  • Configurable Hardware: Optional grippers, wrist force/torque sensors, LiDAR and vision equipment allow the platform to be adapted to different projects.
  • Integrated Protection Features: Emergency-stop hardware, current limits, error monitoring and self-collision safeguards help manage a complex multi-axis system.

Technical Specifications

• Maximum Mobile Speed of 1.5 m/s

• Vertical Body Movement of More than 500 mm

• Arm Reach: 600 mm to the wrist

Max. Size: 600L  690W  1400H (mm)

Max. Size: 23.622L  27.165W  55.118H (in)

Max. Payload: 3 kg (per arm)

Max. Payload: 6.614 lbs (per arm)

Weight: 130 kg

Weight: 286.601 lbs

Max. Locomotion Speed: 1.5 m/s

Max. Locomotion Speed: 4.921 ft/s

Battery Operated? Yes

Locomotion Type: Wheels/Tracks

DOFs per Arm: 7

DOFs per Leg: 6

DOFs (total): 24

Operating Temperature: 40 ℃ Max.

Operating Temperature: 104.000 ℉ Max.

Gripper: 1 DoF per gripper | Stroke: 100 mm | Movement Speed: 40 mm/s | Weight: 530 g | Size: 126 × 55 × 133 mm

Arm Repeatability: Better than ±0.05 mm

Optional Mobility: Mecanum-wheel system for omnidirectional movement

Sensors: LiDAR and 3D perception integration supported; exact sensor package depends on configuration

Software: Python and C++ SDKs, ROS 2 resources and open APIs

Material: Aluminium (Exterior)

Power supply: 48 VDC

Battery: 50 V, 25 Ah, 1,270 Wh

No more specs to show.

Please visit the official website for more accurate, up-to-date, and complete information on this post. Visit Website   External Link Icon Specifications   External Link Icon

Disclaimer: For informational purposes only. The provided data and/or content are intended as references and may contain errors. We appreciate your help in maintaining accuracy; if you spot any errors, please notify us. For the most reliable and up-to-date information, kindly refer to the official website. By using addoobot and its contents and services, you agree to our Terms and Conditions, Privacy Policy, and Cookie Policy.

Thumbs Down Icon Thumbs Up Icon
0 1 2 3 4 5 6 7 8 9 10

Select appealing qualities


 0 Comments

Sign in to comment, reply, and more

No Comments Yet

Be the first to share your thoughts!