Diegetic Graphical User Interfaces and Intuitive Control of Assistive Robots via Eye-gaze
Individuals with tetraplegia and similar forms of paralysis suffer physically and emotionally due to a lack of autonomy. To help regain part of this autonomy, assistive robotic arms have been shown to increase living independence. However, users with paralysis pose unique challenging conditions for the control of these devices. In this article, we present the use of Diegetic Graphical User Interfaces, a novel, intuitive, and computationally inexpensive approach for gaze-controlled interfaces applied to robots. By using symbols paired with fiducial markers, interactive buttons can be defined in the real world which the user can trigger via gaze, and which can be embedded easily into the environment. We apply this system to pilot a 3-degree-of-freedom robotic arm for precision pick-and-place tasks. The interface is placed directly on the robot to allow intuitive and direct interaction, eliminating the need for context-switching between external screens, menus, and the robot. After calibration and a brief habituation period, twenty-one participants from multiple backgrounds, ages and eye-sight conditions completed the Yale-CMU-Berkeley (YCB) Block Pick and Place Protocol to benchmark the system, achieving a mean score of 13.71 out of the maximum 16.00 points. Good usability and user experience were reported (System Usability Score of 75.36) while achieving a low task workload measure (NASA-TLX of 44.76). Results show that users can employ multiple interface elements to perform actions with minimal practice and with a small cognitive load. To our knowledge, this is the first easily reconfigurable screenless system that enables robot control entirely via gaze for Cartesian robot control without the need for eye or face gestures.