Virtual reality (VR) is an exciting technology that enables us to be immersed in virtual world, yet is not adopted by the majority of people. We may feel uncomfortable when using VR goggles because we cannot correctly perceive actual spatial sense in a digital world. In this research, I found that motion sickness can be relieved if compatible haptic feedbacks are provided, however, there was no strong evidence to show a significant correlation between the VR gaming experience and haptic responses.

I participated in the literature review, reading and summarizing the papers about simulator sickness and motion sickness susceptibility. I then designed the experiment tasks and the method to generate compatible haptic feedback, and then conducted the experiment on 20 users. 

My Role:

HCI Researcher

Scope:

Literature review, Experiment design, Data analysis

HCI Lab, Dept. of Industrial Engineering, Tsinghua University

Literature Review

We reviewed paper on VR games and detected that sickness is a common problem among players. Generally, not only VR game players but also FPS or racing game players suffer from nausea. These games intrude our balance system and indicate motions. However, we are actually not, our body stays still in the real world. This may cause disorders in our mind, thus we may feel uncomfortable.

The mechanism of these first person perspective sickness is similar to the one of simulator sickness, but they are not exactly the same. However, approaches that handles simulator sickness may be also helpful for better gaming experience. I then learned many things about simulator sickness and motion sickness, finding their measurement and applicable questionnaires (SSQ & MSSQ).

I wondered that if I give responses related to motion, people will feel relieved to some extend. I then focused on haptic response, which is life-compatible as well as practical in real world.

Experiment Design

Not all VR games are eligible for haptic responses. The first thing I did was finding the appropriate gaming task flow. I evaluated scores of games by their difficulty, controllability, image quality and task time. 

Finally, Project Cars was selected since the researcher can control timing, difficulty and the relative level of motion sickness, and it is friendly to novice players. This game performed perfectly in pilot experiments. InCell and InMind did not have many controls, so they seemed more like VR videos. Time Machine VR was a great context where VR could be applied, however, I found that it was hard to control the experiment time in the pilot runs. 

The most challenging problem was to generate game-compatible and life-compatible haptic responses because the game only provide some vibration on XBOX controller. I used a vibration vast, collaborating with controller, to simulate collisions and bumps on the road in car racing. After discussion, I decided to divide and duplicate low-frequency audio signal from game to drive vibration unit, based on the fact that collision and bump usually create such sounds. Also, when there was a loud sound in the game, I also generated appropriate feedback according to the loudness.

The experiment platform was build by a desktop PC, an Oculus Development Kit 2, a XBOX controller, and most importantly, a vibration vast. The vibration motor is powered by the amplified audio derived from the game original audio track.

The experiment task flow was designed as follows:

• Subject completes questionnaire on motion sickness susceptibility.
• Calibrate Oculus DK2 and familiarize subject with VR vision and control.
• Teach the subject, and pilot race in short track.
• Subject drives on a specific track for about 10 minutes twice.
• Vibration response is switched off during a random run in two rides.
• Subject's times of collision or turnover are recorded as performance data.
• Subject fills simulator sickness questionnaires after every one ride.
• Subject fills the experiment satisfaction questionnaire at last.

Result & Summary

In this experiment, I recruited 20 subjects from various departments in Tsinghua University. I invited 10 male and 10 female, and their age scattered between 20 to 30. Subjects all reported their great experience in VR gaming and their high satisfaction toward our experiment. As for the results, we first cleaned our data to filter some obvious outliners. We analyzed the rest of data as follows:

• Test normality for all data group using Kolmogorov-Smirnov method. All data distributed normally.
• Used paired t-test to find difference between groups (confidence level 0.95).
• Found significant difference in simulator sickness when switching haptic responses, but we could not reject the hypothesis that performance and satisfaction remained the same.

For VR gaming designers, I believe that better image quality, higher frames per second and more vivid effects will help to reduce the sickness during playing. This research suggests that, instead of solely focusing on image, we may use more interactive ways to enhance gaming experience as well as reduce nausea. 

See more projects: