Design of control elements in Virtual Reality - Investigation of factors influencing operating efficiency, user experience and presence

The ergonomic design of control elements in real life has been researched in detail. Various studies exist on the optimal dimensioning, their haptic and acoustic feedback to achieve high control accuracy and user experience. But the development of products is increasingly done with virtual prototypes. Virtual reality (VR) allows these prototypes to be tested in a highly immersive environment. However, the findings from reality cannot be transferred to VR directly. For example, users in VR interact with the prototypes using controllers, which affects haptic feedback. This study investigates how rotary dials and joysticks must be designed and programmed in VR so that control tasks can be performed efficiently and generate a high user experience and perceived presence.In user tests, subjects (n = 25) evaluate the control of a joystick and a rotary dial in VR. In a virtual crane operator's cabin or at a virtual table, the subjects (f = 10, m = 15, age: 24 +- 3) perform four predefined tasks per control element. On two screens in VR, subjects see a vertical bar graph with a scale from 0 to 100 % controlled by the joystick and separately a numerical value between 0 and 100 % which is controlled by the rotary dial. The screens display the task to the subjects, e.g., "Set the value from 0 % to 42 %". According to the method “design of experiments”, 14 factors, such as vibration feedback, acoustic feedback, position of the subject or the sensitivity of the control element are systematically varied on two respectively three levels (e.g., diameter of actuator 40 mm, 80 mm or 110 mm). For each trial, the control accuracy and the time required to complete the task are determined. In addition, the perceived presence is assessed using the Slater-Usoh-Steed-Questionnaire and the user experience is surveyed using the User Experience Questionnaire. The effect of a change of level on the response parameters is investigated using multifactorial ANOVA (α = .05). Linear regression is used to calculate a mathematical relationship between factor and response parameter. These mathematical models are used to calculate which factor values can be used to achieve a high level of control accuracy with a low time requirement and a high level of user experience and perceived presence. The factors angular resolution, inclination, shape of the rotary dial and position of the subject have a highly significant effect (p ≤ .001) on the time required to complete the tasks with the rotary dial. On the control accuracy of the rotary dial, the angular resolution, the VR-controller and the interaction of angular resolution and diameter of the rotary dial have a significant effect. On the user experience, a total of six factors and two interactions have a significant effect. On the perceived presence of the subjects, the VR environment and the diameter of the rotary dial have a significant effect. The calculated optimized design is a rotary dial with vibration feedback, without acoustic feedback, with visualization of a rough knurling, an angular resolution of 10-12 degree/value, a 40 mm diameter and no inclination. Visualization of the hand should be avoided.Sensitivity, size, subject position, VR environment, and the interaction of subject position and VR environment have a significant effect on the time required to perform the control tasks with the joystick. Three factors and one interaction have a significant effect on the control accuracy of the joystick. The interaction of the factors vibration feedback and visualization of the hand has a significant effect on the perceived presence of the subjects. On user experience, nine factors and five interactions have a significant effect. The calculated optimized levels of factors for the joystick are vibration and acoustic feedback, no visual feedback, vertical handle with a height of 20-24 cm, a five-level angular resolution, a maximum deflection angle of +- 15°, a sensitivity of 8 %/sec and a visualization of the hand.The trials show a high degree of scatter. The residuals show outliers in the experiments. These deviations are mainly due to the individual previous experience of the test subjects in handling VR systems. Nevertheless, significant effects could be identified. A screening experimental design was used in this study. In a follow-up study, detailed investigations with a full factorial experimental design must be performed with the significant factors. The factors will be tested at multiple levels and with a significantly increased number of trials to further increase the accuracy of the mathematical models.