A Preliminary Study of Wearable Olfactory-Thermal Feedback for Immersive VR Cultural Heritage
Hanbing Wang, Xiang Li, Kexin Nie, Yunfeng Shu, Xizi Liu, Xiwen Liu, Yingying Liu, Zhixin Cai, Mingjie Zhou, Xianqi Wang, Li Zhao, Chunling Wu, Lie Zhang · 2026 · Extended Abstracts of the 2026 CHI Conference on Human Factors in Computing Systems (CHI EA ’26) · doi:10.1145/3772363.3798804
Summary
This CHI 2026 Extended Abstract reports a preliminary between-subjects user study (N=52) of NeckScents, a lightweight neck-worn wearable that delivers short, pre-authored pulses of scented warm airflow near the face to augment audiovisual VR. The authors position the work within cultural heritage VR, where conveying atmosphere, materiality, and embodied qualities of historical environments is hard to do through audiovisual rendering alone. Rather than rebuilding the visual–audio pipeline, NeckScents acts as a modular sensory add-on: two independent neck-mounted modules sit below the jawline, each housing an absorbent scent pad, a miniature fan, and a low-power heater that can be triggered left, right, or together along an authored VR timeline. The test scenario was a 4-minute guided walkthrough of Kizil Cave 38, a Buddhist heritage site, with four narrative beats (entry → mural viewing → ritual/transition → exit). Participants in the AVOT condition received four cued pulses (2–6 seconds each) using benzoin and lotus floral scents chosen for congruence with Buddhist incense and flower offerings, while the AV control received the same audiovisual content with no scent or warm airflow. The team measured immersion (IEQ), narrative engagement (NES), aesthetic experience (AEQ), awe (AWE-SF), workload (NASA-TLX), and usability (SUS), with Holm–Bonferroni correction applied to the four primary experiential outcomes.
Key findings
Adding olfactory–thermal cues significantly improved three of the four primary experiential outcomes. Immersion (IEQ) rose from 4.08 (AV) to 5.07 (AVOT), F(1,50)=7.27, p=.0095, d=.75. Aesthetic experience (AEQ) rose from 4.02 to 5.20, F(1,50)=12.89, p<.001, d=.996, the largest effect in the study. Awe (AWE-SF) rose from 3.84 to 4.67, F(1,50)=7.07, p=.0105, d=.74. Narrative engagement (NES) trended up (3.76 → 4.42, p=.051) but did not survive Holm–Bonferroni correction. Critically, the secondary measures showed no cost: NASA-TLX workload was statistically unchanged (25.6 vs. 22.2, p=.274) and SUS usability was comparable (60.9 vs. 68.2, p=.098). Internal consistency for all scales was strong (α between .80 and .98). The authors interpret the AEQ and AWE-SF gains as evidence that emotional–aesthetic appraisal is a more sensitive lens than immersion alone for detecting brief multisensory augmentation effects in heritage VR. They explicitly do not claim the specific scent set generalises; rather, the contribution is the event-to-cue authoring template (event → module, duration, intensity) that can be re-authored for other guided heritage narratives.
Relevance
For accessibility practitioners, this paper sits at the intersection of multisensory design, museum and cultural heritage access, and wearable computing. Olfactory and thermal channels are alternative routes to atmosphere and meaning that do not depend on vision or hearing, which makes them potentially valuable for inclusive heritage experiences — though that case is not made directly here. The lightweight, neck-worn form factor and modular add-on philosophy are also useful precedents for designers who want to layer sensory feedback onto existing VR content without forcing a full re-author. Limitations are significant and clearly stated: a single 4-minute scenario, binary present/absent manipulation with no parameter sweeps, no longitudinal or repeated-use data, no testing with disabled or sensory-sensitive participants, and recruitment criteria that excluded people with strong VR intolerance or olfactory impairments. Anyone extending this approach to inclusive design should treat scent intensity, thermal range, and cross-contamination as accessibility variables, not just engineering ones.
Tags: virtual reality · multisensory interaction · olfactory display · thermal feedback · cultural heritage · wearable technology · immersion · aesthetic experience
Standards referenced: NASA-TLX · System Usability Scale (SUS)