AgileX NERO dual 7-DOF arms sync via CAN bus for AI research
Two robot arms mirror each other in real time for embodied AI experiments.
AgileX Robotics has published a comprehensive tutorial on configuring two NERO 7-DOF robotic arms into a master-follower system via CAN bus communication. The setup enables real-time synchronization where the follower arm automatically replicates the master arm’s joint trajectories and end-effector motion. The guide walks through hardware connections (CAN H/L wiring), role assignment for master and follower, parameter saving, and a crucial safety procedure to align both arms before activation to avoid rapid movement and potential collisions. It also notes a software limitation in version 1.1: API access is restricted to commands from the master arm, meaning independent follower state queries are not available.
This dual-arm configuration is designed for researchers and developers working on embodied AI, teleoperation platforms, imitation learning pipelines, and collaborative manipulation. The practical step-by-step instructions allow users to quickly deploy a synchronized robotic system for data collection or experimentation. With dual 7-DOF arms, the platform provides the dexterity needed for complex tasks like assembly, bimanual manipulation, and human-robot interaction studies. The tutorial emphasizes safety and validation procedures, making it accessible to both academic labs and industrial developers building foundation models for robotics.
- Configures two AgileX NERO 7-DOF arms as master and follower via CAN bus for real-time motion replication.
- Safety warning: align arms near home positions before enabling sync to avoid rapid movement and collisions.
- Software v1.1 limits API access to master arm commands; follower state cannot be queried independently.
Why It Matters
Enables affordable, synchronized dual-arm platforms for embodied AI and imitation learning research.