Dual-Arm Robotic Research Platform Based on the OpenArmX Pro Max High Payload, Embodied AI Agility A2

Agility A2 is a high-payload dual-arm robotic research platform built on the OpenArmX Pro Max follower. Each arm offers 7 DOF and up to 5 kg rated / 12 kg peak payload, giving researchers a class-leading bimanual manipulator that is powerful enough for real-world tools and workpieces, yet compact enough to fit on a lab bench. With aluminum-alloy and stainless-steel construction, QDD compliant joints, and a robust 48 V drive system, the platform delivers smooth, human-like motion and reliable long-term operation.

Designed for embodied AI and robot learning, Agility A2 natively supports ROS2, and NVIDIA Isaac, making it easy to collect high-quality teleoperation and scripted data and to deploy imitation learning, reinforcement learning or VLA-based policies on real hardware. The follower can be connected to external leader devices (VR controllers, exoskeletons, custom master arms, etc.) to form a full leader–follower teleoperation system, or driven directly via ROS2 for autonomous experiments. Typical use cases include bimanual manipulation research, contact-rich assembly, human–robot collaboration studies, and university teaching labs that need an open, extensible dual-arm platform.

The robot ships with an integrated industrial PC and a complete software stack. Ubuntu 22.04 with ROS2 Humble is pre-installed, together with an OpenArmX SDK that includes:
motor_manage_openarmx – in-house motor management system for configuration, monitoring and diagnostics
motor_tests_openarmx_com – in-house Python scripts for motor control testing and quick bring-up
openarm_can_openarmx_com – adapted CAN motor driver for OpenArmX
openarm_description – robot model package for simulation and visualization
openarm_ros2 – ROS2 control module for motion control and application integration
openarm_utils – in-house developed teaching / demonstration mode package for quick teleoperation and recording

Use Case:

University & Research Labs – Embodied AI & Robot Learning
For university and institute labs working on embodied AI, robot learning, imitation learning and reinforcement learning. The dual-arm platform with ROS2 is used to collect data and evaluate VLA-based control policies on real bimanual manipulation tasks such as grasping, sorting and assembly.

Hazardous & Specialized Scenarios – Remote Manipulation & Teleoperation
For labs simulating nuclear, chemical or biosafety workflows that require leader–follower teleoperation and remote manipulation. Operators control the dual arms to turn valves, toggle switches, pick up vials and transfer samples, studying teleoperation, safety strategies and contact-rich manipulation without exposing humans to hazardous environments.

Robotics & Industrial R&D – Dual-Arm Assembly & Human–Robot Collaboration
For robotics companies and industrial R&D teams prototyping dual-arm workcells and collaborative robots. The high-payload bimanual manipulator performs contact-rich assembly tasks—plugging connectors, tightening screws, handling tools and clamping or flipping parts—enabling evaluation of impedance control, force control and human–robot collaboration schemes before deployment.

Teaching Labs, Training & Demos – Robotics Education & Hands-On Skills
For university teaching labs, vocational training centers and demo spaces focusing on robotics and automation. The open dual-arm platform supports hands-on courses on robot kinematics, ROS2 control, teleoperation and embodied AI, and can be turned into interactive demonstrations for science museums or trade shows.

Your Purchase Includes:
1x Agility A2 Dual-Arm Robotic Arm（Follower）