The work undertaken so far has focused around our three core research challenges: interaction techniques for confined spaces, social acceptability, and motion sickness.
Interaction in confined spaces
We focused on understanding the design space of the different passenger environments, including cars, aeroplanes, buses and trains. As a first attempt to understand this, we simulated a VR passenger airplane environment to evaluate three different AR-driven virtual display configurations (Horizontal, Vertical, and Focus), exploring their usability, user preferences, and the underlying factors that influenced those preferences. We found that the perception of invading other's personal space significantly influenced preferred layouts in transit contexts. We found that the placement of virtual content was highly influenced by the physical environment, depending on the position of both passengers and physical obstacles. Results from our experiment showed: significant usage of the physical environment to align displays; strong social effects meant avoiding placing displays over other passengers or their belongings; and use of displays for shielding oneself from others. Our findings show the unique challenges posed by the mode of transport and presence of others on the use of AR for mobile productivity in the future.
Social acceptability:
XR headsets have the capability to create virtual content anywhere around the user, disconnecting users from their surroundings. To create solutions for more socially acceptable VR experiences, we first needed to understand what factors and concerns influence social acceptance of VR in different travelling contexts. We found that VR on shorter journeys was less socially acceptable because participants felt that there was insufficient time to enjoy the VR experiences and it was important to follow the route more closely, compared to longer journeys that had a clear final stop, allowing the user to immerse themselves in VR. We also found that accidentally injuring someone, losing one’s belongings or being unable to react if fellow passengers require attention were key barriers to VR adoption. We established that the lack of awareness of immediate objects, one's belongings and other passengers were seen as major safety and comfort concerns for VR use in transit. Therefore, our following work investigated how the visibility of these objects could influence user experience and social acceptance of VR. We designed a user study that brought these elements into Virtual Reality, testing our users’ reactions. We found that other passengers were the most important cue from reality that participants wanted to see, whilst inanimate objects were of less interest. This work sets initial guidelines for more acceptable VR experiences.
Motion sickness
Mitigating motion sickness using neurostimulation is a new era of the in-car use of VR, because it focuses on the enhancement or inhibition of human brain functions themselves without any requirement of sophisticated VR software development. We have a 3-step research plan for our neurostimulation study. Step 1 is to explore the brain biomarkers of VR motion sickness. Step 2 is to implement neurostimulation based on the discovered brain biomarkers, which will be followed by Step 3, an in-car study that aims to verify the developed neurostimulation approach. We have completed steps 1 and 2 and step 3 will begin shortly. For step 1, we confirmed that the left parietal region of our brain is associated with VR motion sickness by using conventional functional segregation methods. On top of this, we found a new brain biomarker (that is, frontoparietal coherence) that is associated with VR motion sickness. This finding not only confirms the involvement of the newly proposed human vestibular network in VR motion sickness but also lays the foundation of the implementation of transcranial alternating current stimulation (tACS), featuring simultaneously modulating multiple regions' brain states. We also found that VR motion sickness impaired cognitive control ability represented by the degree of attentional engagement. Step 2 is now complete. We found that the number of participants who withdrew from the experiments in the non-tACS group (that is, the control group) was three times that of the tACS group.
