Leonidas Deligiannidis

1
Wind and Tactile Feedback Improve User Performance
The VR Scooter

• Leonidas Deligiannidis
• Robert J.K. Jacob

2
VR Scooter

• Simulating a physical scooter.
• Natural and easy travel in a large scale virtual
  environments (user utilizes skills gained in the
  real world).
• Vibrotactile actuators simulate motion speed.
• A fan simulates wind when the scooter is in
  motion.
• Our study shows that wind and tactile feedback
  improved user performance.

3
Traveling in IVE

• Traveling is difficult in general.
• Traveling is as effective as the device utilized.
• Technique/device should be easy to master.
• How many degrees of freedom used at once?

4
Feedback

• Draw users attention at a particular point of
  interest.
• Collision information.
• In general to improve spatial awareness and
  improve level of immersion.
• VR Scooter
• Fan (Variable speed).
• Vibrotactile devices (Variable intensity).
• A real physical device with springs and steering
  mechanism.

5
The VR Scooter Environment
Serial Connections
Vibrotactile device
3D sensor
spring
3D tracker
fan
Relay Controller
Status Interface
Proprietary cable connecting the switches on the
VR scooter to the Input Status Interface.
6
The VR Environment
7
A VR Scooter User
8
Vehicle Based Traveling Technique

• Two Modes
• Fly
• Drive (Default)
• Pointing vs Vehicle-based Traveling techniques

User rotates self, still pointing forward
9
Experiment

• 10 minute training session (devices)
• 10 minutes practice with the VR Scooter VE
  different from the trials VE.
• Two trials per person with fans/tactors and
  without (VR enhanced, VR plain)
• 5 to 10 minutes break between trials.
• 13 subjects total (one couldnt complete), so 12
  subjects total.
• The order was counterbalanced to eliminate
  differences in fatigue and learning.
• 10 out of the 12 subjects performed better in VR
  enhanced (order didnt matter).
• Ground collision enabled.

10
Experiment (Cont.)

• At the end of the experiment
• Q The training session familiarized me with
  using the VR scooter.
• 5-point Likert scale (1. strongly disagree ....
  5. strongly agree)
• The responses show that they were sufficiently
  familiarized with the VR scooter (Mean4.75
  Standard deviation0.45).

11
Experiment (Cont.)

• Travel from start to finish
• Measure Speed and Accuracy
• Subjects told us when they were ready verbally.
• Finally, the subjects filled out a survey
  containing questions about satisfaction with the
  interfaces and the level of immersion they
  presented to the user.

12
Results

• Our hypothesis was that in virtual environments
  subjects would perform better if the level of
  immersion is increased, which was achieved by the
  use of a fan and vibrotactile devices.
• within-subject experimental design.
• Our independent variable traveling type
• (using fan/vibrators and not using
  fan/vibrators).
• Our dependent variable performance
• (time to complete the task getting from the
  start to the finish point).
• Scoring Penalty for inaccuracy.

13
Results (Cont.)

• No scoring penalty, no significant improvement in
  performance
• F(1,11)0.49, p0.498
• With Scoring penalty
• F(1,11)6.88, p0.024

14
Results (Cont.)
15
Results (Cont.)

• Open-ended question Which version do you
  prefer VR plain or VR enhanced and why.
• All twelve subjects specified that overall they
  prefer the VR enhanced.
• experience more convincing, and the environment
  comes to life
• Difficult to figure out their speed of travel in
  VR plain (going through walls, late
  turning/tilting)

16
Results (Cont.)

• Moreover, we asked all subjects if they find the
  utilization of fans and vibrotactile devices
  useful in Virtual Reality and particularly in
  scooter-based traveling techniques.
• The results suggest that the utilization of fans
  and vibrotactile devices are useful in Virtual
  Environments (mean4.83, standard deviation0.39)
  based on a 5-point Likert scale where 1 is
  strongly disagree and 5 is strongly agree.

17
Questions ?