[PDF]

Continuous Eye Tracking Framework For Manual Input Comparison With High-Level Telepresence Robots


Pun Ching Desiree Yan

03/10/2024

Supervised by Juan Hernandez Vega; Moderated by Usashi Chatterjee

Human-robot interaction (HRI) and human-robot collaboration (HRC) are relevant areas of robotics that explore the necessary requirements that allow robots to work alongside humans. One key aspect of such collaboration is establishing an efficient transmission of information between robots and humans. A recent and novel approach to this challenge has been the use of virtual reality (VR). Current gaze-control layout for teleoperation often suffers from various challenges such as ambitious layout and reliance on user memory, which leads to high cognitive load. This project aims to design a continuous gaze tracking input for teleoperation in the Unity game engine to reduce these issues and use it as a baseline to compare telepresence robots with high-level controls. After that, a user study was conducted to compare continuous system with discrete teleoperation and semi-autonomous navigation. Unity application would then deployed onto the Meta Qest Pro's headset for users studies so the users can control the teleprescene robot in either low or high level commands in random order,users could then navigate to a specific target inside the virtual museum. The study will evaluate the purposed framework could be feasible as a baseline base on quantitative data and qualitative data gather from the study.


Final Report (03/10/2024) [Zip Archive]

Publication Form