Collaborative design activities, such as design critique and brainstorming, typically center around physical artifacts. These activities often require the use of complex non-verbal cues, including head and body movement, facial expressions, and pointing gestures. In co-located settings, peripheral awareness enables participants to understand each other’s locus of attention with ease. However, these spatial cues are mostly lost while using videoconferencing tools. Although previous efforts have proposed video and MR-based systems to digitally represent individuals, the physical embodiment of presence can offer substantial benefits, especially when collaboration involves physical artifacts. Telepresence robots have been implemented for this purpose, but they often require explicit control interfaces like a mouse or keyboard for navigation. This requirement is not well-suited for the cognitively demanding design tasks, as it divides users from their primary tasks, reducing collaboration effectiveness. To bridge this gap, my research investigates the design of implicitly controlled telepresence robots that facilitate collaboration around physical artifacts. These systems are designed to automatically translate head, body, and facial movements into robotic actions to match the user’s intentions. The focus is on physically representing non-verbal cues, while ensuring users can concentrate on their primary design tasks with minimal distraction. To explore this design space, I have developed a series of embodied robotic systems, each designed to tackle increasingly more complex scenarios. This includes integration of a VR interface to accommodate asymmetric settings, enabling a remote collaborator to navigate and collaborate in multiple rooms or buildings from a small space.