Bimanual Manipulation Platform Products

Almond Axol

Almond

Stationary dual-arm robot for physical AI: 2x 7-DoF arms, 860 mm reach, 4 kg nominal (6.5 kg peak) payload per arm, 1 mm repeatability, 19.5 kg. MIT-licensed Python SDK with WebXR VR teleoperation and LeRobot bindings. From $7,999.

RealMan RMC-DA-LIII Dual-Arm Lift

RealMan Robotics

A mobile dual-arm platform from RealMan built for wired teleoperation, dual-arm data collection, and AI/imitation-learning model training. Two 6-DoF RM65-B-V arms (5 kg payload each) mount on a 900 mm vertical lift atop an autonomous mobile base, giving a 0–2400 mm working height. Widely used in Chinese labs for bimanual teleop and robot-learning data pipelines.

Dual UR5e

Universal Robots

A stationary bimanual research platform composed of two dual-mounted Universal Robots UR5e 6-DoF collaborative arms. Not a vendor-sold bimanual SKU but a recognized industrial research configuration, valued for its rugged industrial duty cycle, wide software support, and RTDE real-time control interface. Each UR5e offers 850 mm reach, 5 kg payload, and ±0.03 mm repeatability, making the pair well suited to applied/industrial learning and tactile/force manipulation experiments.

WowRobo TwinArm

WowRobo Robotics

Low-cost commercialized dual-arm leader-follower kit from WowRobo Robotics (China), built to the open-source Hugging Face LeRobot Koch v1.1 hardware spec. Each kit ships one 6-DoF leader arm and one 6-DoF follower arm (12 DYNAMIXEL servos total) with a soft-and-hard-material gripper, targeting imitation-learning, teleoperation, and embodied-AI research in the LeRobot ecosystem.

Elephant Robotics myBuddy 280

Elephant Robotics

myBuddy 280 is Elephant Robotics' integrated dual-arm desktop collaborative robot: two 6-DoF arms mounted on a shared rotating waist for 13 total DoF, powered by an onboard Raspberry Pi (280 Pi variant). Marketed as the most cost-effective Raspberry-Pi dual-arm robot, it targets education, teaching, and basic bimanual imitation and dual-end-effector experiments. Each arm offers 250 g payload, 280 mm reach and ±0.5 mm repeatability, with two HD cameras and a 7-inch interactive touchscreen.

Seeed reBot Arm B601

Seeed Studio

Open-source, low-cost 6-DoF robotic arm from Seeed Studio for embodied-AI data collection and imitation learning. Each B601 follower arm is teleoperated by a reBot Arm 102 leader arm; the system supports both single-arm and bimanual (dual-arm) leader-follower setups. Deep LeRobot integration plus ROS 1/2, NVIDIA Isaac Sim and Pinocchio support. Available in DM (Damiao CAN motor) and RS (RoboStride motor) variants.

StarAI Star Arm Bimanual Kit (Violin + Viola/Cello)

Fashionstar

A LeRobot-native, stationary bimanual leader-follower kit from Fashionstar (sold via Seeed Studio). Two Violin leader arms teleoperate two follower arms (Viola, 300g payload / 470mm reach, or Cello, 750g payload / 670mm reach), all 6-DoF + parallel gripper with brushless servos. Ships pre-assembled with explicit LeRobot bimanual configs, plus ROS 2 and MoveIt2 support, targeting imitation-learning and pick-and-place data collection.

SO-ARM101 Dual (bimanual)

The Robot Studio

Ultra-low-cost, fully open-source (Apache-2.0) bimanual leader-follower platform built from four 3D-printed 6-DoF SO-ARM101 arms driven by Feetech STS3215 servos. Designed by The Robot Studio with Hugging Face, it is the de-facto standard bimanual rig in the LeRobot community for ACT/imitation-learning data collection, with kits available from Seeed Studio and Waveshare for a few hundred dollars.

ALPHA-α (Bi-ACT)

The University of Osaka

ALPHA-α is a low-cost bimanual leader-follower teleoperation platform built from four 6-DOF OpenMANIPULATOR SARA arms (two leaders, two followers). It uses four-channel bilateral control to teleoperate with both position and force feedback, enabling force-aware imitation learning (Bi-ACT) for everyday two-handed manipulation. Developed at The University of Osaka as an open, roughly-half-the-cost alternative to ALOHA-class rigs.

Galaxea R1 Pro

Galaxea Dynamics

Higher-end variant of Galaxea Dynamics' R1 wheeled humanoid: a dual-arm mobile manipulation platform with two 7-DoF force-controlled arms, a 4-DoF torso, and a 6-DoF wheeled chassis, built for embodied-AI and VLA data collection.

Galaxea R1 Lite

Galaxea Dynamics

Galaxea R1 Lite is a lower-cost, wheeled dual-arm mobile manipulation platform from Galaxea Dynamics. It pairs two Galaxea A1-class (A1X) 6-DoF arms with G1 grippers on a 3-DoF torso and an omnidirectional wheeled base, targeting whole-body household manipulation and vision-language-action (VLA) research. Isomorphic leader-follower teleoperation and a LeRobot-format Open-World Dataset make it a data-collection and robot-learning platform out of the box.

UFACTORY xArm Dual

UFACTORY

A stationary bimanual manipulation platform composed of two UFACTORY xArm collaborative arms (2x 6-DoF xArm 6, or 2x 7-DoF xArm 7) mounted on a shared frame. Positioned as an affordable dual-arm rig for imitation learning and dual-arm manipulation research, with Python/ROS control and UMI-style teleoperation (FastUMI) for demonstration collection.

Dual Franka FR3

Franka Robotics

A stationary bimanual research platform composed of two 7-DoF Franka Research 3 (FR3) torque-controlled arms mounted on a shared frame. A recognized robot-learning configuration for force-aware imitation and precise teleoperation, built on Franka's torque sensing, impedance/compliance control, and the Franka Control Interface (FCI).

ABB YuMi (IRB 14000)

ABB

ABB's YuMi (IRB 14000) is an inherently-safe, integrated dual-arm collaborative robot built for small-parts assembly. Each of its two 7-DoF arms handles a 0.5 kg payload with 0.02 mm position repeatability, and lead-through (hand-guiding) programming makes it approachable for teleoperated demonstration and occasional bimanual manipulation research.

Mobile ALOHA

Stanford University

Mobile ALOHA is Stanford's low-cost whole-body bimanual teleoperation and imitation-learning reference design: an ALOHA-style leader-follower dual-arm rig (2 leader + 2 follower 6-DOF arms) mounted on a wheeled mobile base, letting a single operator puppeteer both arms and drive the base simultaneously to collect demonstrations for mobile manipulation tasks (cooking, dishwashing, taking an elevator). Co-training with the earlier stationary ALOHA dataset raises success rates by up to ~90%. Hardware BOM and ACT training code are open-source.

ALOHA 2

Trossen Robotics

ALOHA 2 is an enhanced low-cost reference design for bimanual teleoperation and imitation-learning data collection, sold as kits by Trossen Robotics. A stationary tabletop rig pairs two low-cost leader arms with two follower arms (6-DOF each) for leader-follower puppeteering, adding low-friction rail grippers, passive gravity compensation, and Intel RealSense D405 depth cameras. It ships with a high-fidelity MuJoCo model and open-source hardware, and underpins large-scale imitation/VLA datasets and Google DeepMind's Gemini Robotics demos.

ALOHA (original)

Stanford University

ALOHA (A Low-cost Open-source Hardware system for Bimanual Teleoperation) is a stationary tabletop leader-follower rig from Stanford (Tony Z. Zhao et al.), built for fine-grained imitation learning. Two ViperX 300 6DOF follower arms are puppeteered by two WidowX leader arms; a human operates the leaders and the followers mirror the motion while multi-view cameras record demonstrations. Paired with the ACT (Action Chunking with Transformers) policy, it became the most-cited low-cost bimanual data-collection platform from 2023 onward. The design is fully open-source (3D-print parts, assembly tutorial, training code) and is also sold as a kit by Trossen Robotics for roughly a $20k build budget.

TRLC-DK1-X

The Robot Learning Company

The TRLC-DK1-X is a fixed benchtop bimanual robot-learning developer kit from The Robot Learning Company (Germany). It ships two leader-follower arm pairs (four arms total) of TRLC's own open-source DK1 arm design in the SO-100/GELLO/Koch lineage, plus three USB cameras, targeting imitation-learning workflows with native HuggingFace LeRobot integration. Each follower arm offers 700 mm reach and 1 kg payload. The full hardware and software stack (CAD in STEP, URDF, code) is released under Apache-2.0. Sold direct at $6,999 via public Stripe checkout.

OpenArm 2

Enactic, Inc.

OpenArm 2 is a fully open-source bimanual manipulation platform: two 7-DoF quasi-direct-drive arms (Enactic's own DAMIAO-motor design) built for bilateral force-feedback leader-follower teleoperation and imitation-learning data collection. Each gripper has a built-in RGB camera; the fixed/benchtop system integrates with ROS 2 / MoveIt2 / Isaac Lab and is listed as a supported LeRobot robot. Sold factory-assembled by WowRobo (from ~$6,000) and by RT Corporation (by quotation).

Galaxea R1

Galaxea Dynamics

Galaxea R1 is a full-size wheeled dual-arm mobile bimanual manipulation platform built for embodied-AI research and data collection. It pairs two 6-DoF Galaxea A1 arms with G1 parallel grippers on a 4-DoF torso and an omnidirectional three-wheel steering base, giving a 0-2 m vertical working range and a 70 cm horizontal reach per arm (5 kg payload at 0.5 m). An 8-camera vision suite (1 head, 2 wrist, 5 chassis) plus a 360 degree LiDAR feed up to ~550 TOPS of onboard compute. It supports isomorphic and VR teleoperation, ROS 1/ROS 2, and ships with Galaxea's open-source GalaxeaVLA models and an Open-World Dataset released in LeRobot and RLDS format.

AgileX Cobot Magic

AGILEX ROBOTICS

Open-source bimanual mobile manipulation platform based on the Stanford Mobile ALOHA architecture. A leader-follower rig of four 6-DoF AgileX PiPER arms (2 leader + 2 follower) mounted on an AgileX Tracer mobile base, with three Orbbec Dabai RGB-D cameras, industrial-grade onboard compute, and the Mobile ALOHA imitation-learning software stack for teleoperated demonstration collection and dual-arm policy learning.

Trossen Mobile AI

Trossen Robotics

Trossen Robotics Mobile AI is a wheeled bimanual manipulation platform and the successor to Mobile ALOHA. It combines two Leader-Follower WidowX AI arm pairs (four 6-DOF arms total) with three Intel RealSense D405 depth cameras on a mobile base, targeting imitation-learning and teleoperation research with a LeRobot / ROS 2 / Isaac / MuJoCo / Gazebo software stack.

Trossen Robotics Stationary AI

Trossen Robotics

Trossen Robotics Stationary AI is a stationary bimanual imitation-learning workstation and the current successor to ALOHA Stationary. It ships as two leader-follower WidowX AI arm pairs (four 6-DoF arms total) with four Intel RealSense D405 RGB-D cameras, wired for leader-follower teleoperation data collection and policy training across LeRobot, Hugging Face, ROS2, NVIDIA Isaac, MuJoCo, Gazebo, and OpenPI/pi0.