ReMotion: Supporting Remote Collaboration in Open Space with Automatic Robotic Embodiment

Document Title

ReMotion: Supporting Remote Collaboration in Open Space with Automatic Robotic Embodiment

Document Information

• Subject Area: Human-Computer Interaction and Remote Collaboration Design
• Keywords: Robotic Embodiment, Telepresence Robots, Remote Collaboration, Open Space, Automatic Control, Video Conferencing, Peripheral Awareness, Joint Attention

Research Background and Problem

• Problems and Challenges:

  1. Design activities often require participants to move within open spaces, understand each other's focus of attention and body posture, which traditional remote conferencing tools struggle to capture.
  2. While mobile telepresence robots can improve the sense of presence in remote collaboration, manual control imposes a high cognitive load, especially in complex open spaces.
  3. Existing systems are typically limited to static scenarios such as meeting tables, making them unsuitable for larger rooms and dynamic collaborative spaces.

• Research Significance: Achieving natural and seamless spatial movement and peripheral awareness is crucial in remote design activities. This directly impacts team collaboration efficiency, sense of participation, and understanding of others' attention.

• Research Motivation and Related Work:

  • Previous studies have focused on improving remote collaboration through video conferencing and robotic proxies, but these are mainly confined to static, small-scale scenarios (e.g., meeting tables).
  • While robotic systems with automatic motion control have been preliminarily explored, achieving embodiment in broader open-space tasks remains a challenge.
  • Peripheral awareness and body movements provide critical information for dynamic collaboration transitions, necessitating the integration of these features into remote collaboration systems.

Solution

• Method or Solution:

  • The ReMotion system is a novel mobile robotic platform capable of automatically mimicking the body and facial movements of remote collaborators, embodying them in physical space.
  • Kinect Azure is used to capture the user's body position and orientation, while the NeckFace wearable device tracks head direction and facial expressions.
  • An omnidirectional drive platform enables precise movement across a wide range, offering greater flexibility in replicating complex human motions compared to traditional differential drive systems.
  • The system employs a symmetrical configuration, ensuring that both remote and local users can interpret each other's attention and intentions through the same framework.

• Innovations:

  1. Combines an omnidirectional drive system with wearable facial tracking devices to enhance synchronization of movement and facial expressions.
  2. Implements automatic control, significantly reducing the cognitive load on remote users.
  3. Introduces a low-degree-of-freedom robotic arm to enhance gesture-based pointing while addressing laser pointer accuracy issues.

• Implementation Steps:

  1. Body Motion Capture and Mapping: Kinect captures user body movements, which are rendered in real-time using the omnidirectional platform.
  2. Head and Facial Rendering: NeckFace extracts head orientation and facial expression data, which are displayed using a motion display.
  3. Pointing Representation: A low-degree-of-freedom robotic arm is added to simulate hand-pointing gestures.
  4. Hardware and Software Integration: Unity game engine and Python scripts are combined for data processing and robot control.

Research Outcomes

• Specific Results:

  1. The ReMotion system significantly enhances "co-presence" and "behavioral interdependence" in remote collaboration.
  2. Participants in experiments were able to dynamically and seamlessly understand the remote collaborator's attention and position through the robot.
  3. Experiments showed that ReMotion provided higher shared attention time while reducing the frequency of attention shifts.

• Advantages Over Existing Solutions:

  • Outperforms wheeled robots and traditional video conferencing tools in supporting dynamic tasks within open spaces.
  • Automatic control reduces operational demands on remote users, enhancing their sense of participation and focus.

• Experimental or Evaluation Results:

  • Shared Attention Time: The average shared attention time under the ReMotion condition was significantly higher than with traditional video systems.
  • User Feedback: Most users preferred the ReMotion system, describing it as offering a more "immersive" experience.

• Limitations and Future Directions:

  1. Limitations:
     • The system relies heavily on symmetrical environment assumptions, making it challenging to adapt to diverse real-world configurations.
     • Dynamic head motion rendering is not fully implemented, leaving user needs for facial expression representation unmet.
     • The robot may experience some vibration during rapid movements, affecting stability.
  2. Future Directions:
     • Introduce adjustable height mechanisms to support more complex interaction scenarios (e.g., transitions between sitting and standing).
     • Improve facial and gaze tracking accuracy to enhance non-verbal communication expressiveness.
     • Explore applications in asymmetric scenarios by implementing dynamic remapping for spatial coordination.
     • Add physical manipulation capabilities to accommodate more diverse design task requirements (e.g., remote teaching or demonstrations).

Conclusion

ReMotion offers a novel approach to remote collaboration by enabling dynamic interaction across spaces through an automatic mobile proxy. It not only enhances users' peripheral awareness but also effectively supports joint attention in shared open spaces. Its potential holds significant value for future applications across various fields, particularly in education, design, and cross-regional collaborative projects.