[Paper Review] Defining Preferred and Natural Robot Motions in Immersive Telepresence from a First-Person Perspective

Katherine J. Mimnaugh, Markku Suomalainen+4 more|arXiv (Cornell University)|Feb 25, 2021

Virtual Reality Applications and Impacts13 references4 citations

TL;DR

This study investigates how autonomous robot motion affects user comfort and perceived naturalness in immersive telepresence via first-person virtual reality. Using a user study with 36 participants, it finds that path comfort and subjective naturalness strongly influence preference, with turn quality and proximity to objects being key factors—leading to recommendations for motion planning that prioritize both physiological comfort and human-like motion perception in telepresence robots.

ABSTRACT

This paper presents some early work and future plans regarding how the autonomous motions of a telepresence robot affect a person embodied in the robot through a head-mounted display. We consider the preferences, comfort, and the perceived naturalness of aspects of piecewise linear paths compared to the same aspects on a smooth path. In a user study, thirty-six subjects (eighteen females) watched panoramic videos of three different paths through a simulated museum in virtual reality and responded to questionnaires regarding each path. We found that comfort had a strong effect on path preference, and that the subjective feeling of naturalness also had a strong effect on path preference, even though people consider different things as natural. We describe a categorization of the responses regarding the naturalness of the robot's motion and provide a recommendation on how this can be applied more broadly. Although immersive robotic telepresence is increasingly being used for remote education, clinical care, and to assist people with disabilities or mobility complications, the full potential of this technology is limited by issues related to user experience. Our work addresses these shortcomings and will enable the future personalization of telepresence experiences for the improvement of overall remote communication and the enhancement of the feeling of presence in a remote location.

Motivation & Objective

• To understand how autonomous robot motion affects user experience in immersive telepresence via head-mounted displays (HMDs).
• To identify which motion characteristics—such as path smoothness, turn quality, and proximity to obstacles—impact user comfort and perceived naturalness.
• To develop a framework for categorizing natural motion from the first-person user perspective, accounting for individual and contextual variation.
• To inform future motion planning in telepresence robots by grounding design in user-centered evaluation.
• To address limitations in current telepresence systems, including VR sickness and unnatural motion, through user-informed design.

Proposed method

• Conducted a user study with 36 participants (18 females) viewing panoramic VR videos of three distinct robot paths through a simulated museum.
• Collected subjective feedback via questionnaires on path preference, comfort, and perceived naturalness of motion.
• Analyzed responses to categorize perceptions of naturalness along two dimensions: time frame (immediate vs. holistic path) and generality (context-specific vs. general behavior).
• Compared piecewise linear paths with smooth paths to assess differences in user experience, focusing on turning behavior and proximity to walls/objects.
• Used qualitative analysis to identify recurring themes in user responses about how they would have moved through the museum had they been in control.
• Planned follow-up experiments with an empty museum environment to isolate motion effects from visual context.

Experimental results

Research questions

• RQ1 How do different robot motion profiles—specifically piecewise linear versus smooth paths—affect user comfort in immersive telepresence?
• RQ2 To what extent does the perceived naturalness of robot motion influence path preference in a first-person VR setting?
• RQ3 How do users conceptualize 'natural' motion in telepresence, and what dimensions underlie their perception?
• RQ4 What role do turn quality, speed, and proximity to obstacles play in shaping user experience and preference?
• RQ5 How does the presence of visual context (e.g., museum artwork) affect judgments of comfort and naturalness in robot motion?

Key findings

• Comfort had a strong and significant effect on path preference, with users consistently favoring paths that minimized discomfort.
• Perceived naturalness also strongly influenced preference, even though users defined 'natural' in diverse and individualized ways.
• Turn quality and proximity to walls or objects were the most frequently cited factors in judgments of naturalness, with sharp turns and close proximity being perceived as unnatural.
• Users often described natural motion in terms of how it aligned with their own anticipated behavior—such as slower speeds and deliberate turns when viewing exhibits—indicating context-dependent naturalness.
• The path that allowed closer inspection of museum items was rated as most natural, suggesting that motion that supports observational behavior is perceived as more human-like.
• There is no universal definition of natural motion; instead, naturalness is context-dependent and varies across individuals, making user evaluation essential for motion design.