Teleoperation

Master-Follower Teleoperation

Master-follower teleoperation, also called leader-follower teleoperation, is a control scheme in which an operator moves a kinematically similar leader device and a follower robot tracks its joint positions in real time. Because mapping is joint-to-joint, it is intuitive and low-latency, requiring no inverse kinematics. ALOHA and the Koch and SO-100 arm pairs use this scheme with low-cost leader arms to collect fine bimanual manipulation demonstrations.

Why it matters for physical AI

Joint-space puppeteering proved to be the highest-bandwidth affordable way to capture contact-rich demonstrations, and it underlies most of the bimanual manipulation datasets now used to train imitation policies.