Dual 7-DoF arms with 70 cm workspace reach and spherical human-like wrist design
10 kg combined arm payload with 5 kg per arm extended and ±0.5 mm repeat positioning accuracy
VR teleoperation via Oculus Quest 3 with 100 ms latency and active safety boundary protection
Intel Core i7-1165G7 AI computing module with 2 TB storage and ROS1/ROS2 support
Native VLA platform with ACT/Pi 0.5 classic models and 10,000+ open-sourced real-world datasets
Python and C++ SDK with NVIDIA Isaac Sim, MuJoCo, and Gazebo simulator compatibility
LimX Dynamics TRON 2 EDU Dual-Arm Embodied Robot is a high-performance dual-arm manipulation platform developed by LimX Dynamics, purpose-built for Vision-Language-Action research, teleoperated data collection, and bimanual task development. Each arm features seven degrees of freedom with a spherical wrist design that enables wide-range articulation across the full 70 cm workspace, producing natural postures and precise end-effector placement for complex manipulation tasks. Active safety boundary protection covers the head, body, stand, and desktop zones, while a dual redundant power design keeps the arm modules in a locked safe state in the event of unexpected power loss. Suited for academic research labs, AI development teams, and robotics engineering programs focused on manipulation, dexterous control, and embodied learning.
Perception is handled through RGB-D cameras positioned at the head and wrist, providing full-field visual coverage of the manipulation workspace for accurate closed-loop feedback during both autonomous and teleoperated operation. VR-based dual-arm control via Oculus Quest 3 delivers low-latency remote operation, enabling efficient high-quality trajectory demonstration recording for imitation learning workflows. The integrated VLA development platform consolidates data collection, annotation, model training, and inference into a single streamlined interface, with a large library of open-sourced datasets and classic models including ACT and Pi 0.5 available immediately. With a fully open SDK, ROS1/ROS2 compatibility, and support for NVIDIA Isaac Sim, MuJoCo, and Gazebo, this autonomous robot platform is designed to serve researchers, university laboratories, and development organizations accelerating progress in embodied AI and dexterous manipulation.
Why Buy From RobotShop
Official Authorized Distributor
Genuine products with manufacturer-approved warranty coverage.
Locally Managed Logistics
Customs, inspection, and prep handled locally before shipment.
Single Point of Contact
Sales, support, and returns managed by one team.
FAQ & Knowledge Base
Docs, tutorials, and community answers to move faster.
DoF per Arm
7
Dual-Arm Payload
10 kg
Arm Reach
70 cm
Teleoperation Latency
100 ms
Repeatability
±0.5 mm
Model Features
Precision dual-arm mechanics built for high-payload manipulation, real-time teleoperation, and full-workspace perception.
Arm
Performance→
Arm Performance
Manipulation
Arm DoF: 7 per arm
Combined Payload: 10 kg (5 kg per arm extended)
Max End-Effector Speed: 5 m/s
Max Acceleration: 36 m/s²
Repeatability: ±0.5 mm
End
Effector→
End Effector
Gripper
Type: Gripper (Dexterous Hand not included)
Gripping Force: 20 N
Max Gripper Stroke: 85 mm
Teleoperation→
Teleoperation
VR Control
Latency: 100 ms
VR Device: Oculus Quest 3
Safety Boundaries: Head, body, stand, desktop
Power Design: Dual redundant, anti-drop on power loss
Capabilities
A dedicated dual-arm research platform combining precision manipulation hardware with a complete AI development ecosystem for VLA model training and dexterous task execution.
High-Precision Dual-Arm Control
Each 7-DoF arm offers wide-range articulation with a spherical wrist for natural posture generation across a 70 cm workspace. Industry-leading motion algorithms deliver stable, balanced control during complex bimanual and single-arm manipulation tasks.
Native VLA Development Platform
The integrated platform consolidates data collection, cleaning, annotation, model training, and inference into a single unified interface. Classic models including ACT and Pi 0.5 are preloaded alongside 10,000+ open-sourced real-world datasets for immediate use.
Multi-Zone RGB-D Perception
Cameras at the head and wrist positions deliver full visual coverage of the manipulation workspace for closed-loop feedback during both teleoperated and autonomous task execution. The full-field arrangement eliminates blind spots across the complete arm range of motion.
Open Developer Ecosystem
Python and C++ SDKs provide full high-level and low-level control access with ROS1 and ROS2 compatibility and clear URDF models optimized for Sim2Real performance. Support for NVIDIA Isaac Sim, MuJoCo, and Gazebo enables simulation-first development workflows.
Use Cases & Application Scenarios
From VLA model training and teleoperated data collection to academic manipulation research and shared autonomy experiments, the Dual-Arm configuration serves a broad range of research and development workflows.
VLA Research & Model Training
Research teams use the dual-arm platform to build and evaluate Vision-Language-Action models on standardized manipulation benchmarks. The native VLA platform and preloaded classic models provide an immediate starting point, significantly reducing experiment setup time.
Teleoperated Demonstration Collection
Using Oculus Quest 3, operators record high-quality trajectory demonstrations across a wide range of bimanual and single-arm manipulation tasks for imitation learning. Low teleoperation latency and active safety boundaries make extended data collection sessions both practical and safe.
Academic Manipulation Research
University robotics departments use the platform for controlled studies in grasping, placement, sorting, and bimanual coordination tasks. Comprehensive tutorial resources and an open SDK lower the barrier for new lab members joining the platform.
Shared Autonomy Experiments
The platform serves researchers studying how AI-generated actions and human operator inputs can be blended in real time for cooperative task execution. Wrist and head RGB-D sensors provide the spatial data needed to evaluate interaction dynamics in structured workspace settings.
LimX Dynamics TRON 2 EDU Dual-Arm Embodied Robot
Model: TRON2-A-DA-EDU
•
TRON2 Main Body (includes two upper arms)
•
TRON2 Lower Arm Ends ×2 (includes two grippers)
•
TRON2 Head Unit (vision & audio sensing head)
•
Wrist D405 Camera Set ×2 (includes cameras, mounts, and data cables)
•
TRON2 Battery & Charging Dock (includes power adapter)
•
Remote Controller
•
Oculus Quest 3 Kit (for VR-based data collection and teleoperation)
•
Dev Expansion Module
•
Dual-Arm Tool Kit (specific assembly tools for arms)
•
User Manual & Certificate of Quality
•
Fixed Stand & Fixed Base (DA model, boxed separately)
Package contents vary by configuration. Mobile (DAU) variant includes Lifting Stand & Mobile Chassis in place of Fixed Stand & Base. Confirm final box contents with sales prior to ordering.
LimX Dynamics TRON 2 EDU Dual-Arm Embodied Robot
Model: TRON2-A-D-EDU
Arm Reach
•
70 cm per arm
Max Gripper Stroke
•
85 mm (3.35 in)
DoF per Arm
•
7
Material
•
Aluminum Alloy and Plastic
Overall body dimensions depend on the mounting stand configuration. Values shown are from the manufacturer's reference.