Measuring Detection and Habituation of Olfactory Stimuli in Virtual Reality for Improved Immersion

Abstract: Virtual reality (VR) is a powerful tool that allows humans to interact with systems at scale, often with more feasibility and/or accessibility. However, the usefulness of these digital counterparts can be limited by the immersiveness of the experience, especially when the human-system interaction is integral for the use case (e.g., VR for training or gaming). Therefore, increasing a users’ sense of presence can improve the utility of VR. Previous literature suggests that integrating olfactory may be beneficial towards this. However, gaps exist in formalizing the integration and deployment of olfactory in VR. The objective of this study is to investigate the effectiveness of a wearable odor device for detecting scent stimuli in VR. Our primary goal is to determine the optimal parameters under which a participant can accurately detect a scent stimulus. We seek to answer the following objectives: 1) how do various levels of scent intensity and duration effect scent detection, and 2) does habituation of intensity and pleasantness of scent occur after prolonged exposure.A 20-minute VR study (N=34) using an Oculus Quest 2 and odor attachment was conducted; during which participants were exposed to two scents across various scent intensities and delays. The study used a 5 (intensity: 105, 150, 225, 300, 600 ms; within-subject) x 3 (delay: 15, 30, 60 seconds; within-subject) x 2 (scent: pleasant (smoke), unpleasant (body odor); within-subject) factorial design. Pleasant and unpleasant scents were selected using a pilot study. Participants clicked a button using the Oculus controller when they detected a scent and rated the scent intensity and pleasantness each on a 7-point Likert scale.A binary logistic mixed model was used to predict scent detection (hit/miss). There was a significant effect (p < .05) on detection based on intensity, age, and interaction of scent with intensity. Detection for 600 ms intensity was 4.41 times more likely than intensity at 105 ms. There was a decreased likelihood of detecting the pleasant scent at 600 ms, suggesting that the unpleasant scent was even more detectable at this high intensity. Individuals ≤ 35 years were 2.17 times more likely to detect a scent than those over 35.ANOVAs were performed to assess Likert scale ratings of scent intensity and pleasantness, as a measure of habituation. Results indicate no effect of order or delay, but within-subject effects of scent and intensity. Where unpleasant was rated more intense (p < .001) and more unpleasant (p < .001) compared to pleasant. Additionally, higher intensity dispersions were self-reported as more intensity (p = .003).Our framework and associated conclusions can guide olfactory integration into VR environments for improved human interaction. The results suggest that the effectiveness of a scent may vary depending on the strength of the intensity and the user's age. Certain scents may be more potent at lower intensities, while others may require higher intensities to produce a significant [detectable] effect. The use of multiple sensory integration into virtual training and/or gaming simulations could have a positive effect on knowledge outcomes, skill acquisition, and enjoyment.

Keywords: Perception, Extended Reality, Detection, Olfactory, Simulation and Gaming

DOI: 10.54941/ahfe1005001

Cite this paper:

​