Abstract

Telepresence systems enable remote collaboration through physical robots or virtual avatars, yet prior comparisons have focused on technical capabilities and affordances from remote users. This overlooks how local collaborators interpret and respond to embodied actions. To investigate these differences under comparable conditions, the authors developed a telepresence system that supports functionally equivalent collaboration through either a robotic or avatar embodiment and conducted a within-subjects study in which participants collaborated with a remote partner embodied across various tasks, including navigation, gesture, object referencing, and digital tool use. Despite parity of system functions and tasks, participants consistently interpreted and responded differently to the two embodiments.