Sniff It: A Mobile Companion for Olfactory Training and Everyday Smell Journaling

Background: Olfactory dysfunction can substantially affect quality of life and may persist after viral infections, including SARS-CoV-2 infection. Olfactory training is a commonly used non-pharmacological approach in which individuals repeatedly smell a small set of familiar odors over an extended period. Because training is typically performed independently at home, adherence, documentation, and integration of training into everyday life remain practical challenges. Objective: We developed Sniff It, an iOS application designed to support structured olfactory training while extending the training experience into longitudinal self-observation and everyday smell awareness. System design: Sniff It combines guided olfactory-training sessions with reminders, longitudinal statistics, gamification, and user-controlled data export. A personal scent library allows users to organize familiar training scents and record their own impressions before viewing reference sensory vocabulary. Scent Moments provide a lightweight journal for recording olfactory experiences encountered in everyday life. An optional Smart Scent Capture feature uses computer-vision and text-recognition technologies, with optional on-device Foundation Models assistance on supported devices, to identify visible objects or product labels. When a reliable catalog match is available, candidate results are mapped to a curated scent-reference catalog; otherwise, users may confirm or edit the detected object name and add it as a custom scent. The system does not attempt to infer odor perception from images, and users confirm suggested matches before they are saved. Safety-related decisions and health interpretation remain deterministic rather than being delegated to generative AI. Personal app data are stored locally on-device and are not automatically transmitted to a research server. Planned evaluation: Sniff It Version 1 implements the described training, journaling, personal scent tracking, and data-management architecture. The application has not yet undergone prospective evaluation of adherence or clinical outcomes, and no clinical effectiveness claims are made. A first prospective evaluation is planned to examine usability, acceptability, and naturalistic use of Sniff It. Following an initial structured usability assessment, participants will indicate whether and how they intend to continue using the application, including structured olfactory training, Scent Moments, both, or another mode of use. Participants will then use the application according to their own preferences over a four-week observation period. Follow-up assessment will examine actual use, perceived usefulness, barriers to continued engagement, training adherence among participants who initially intended to train, and acceptance of journaling, gamification, and mascot-based features. The final protocol, sample size, outcomes, and statistical analysis plan will be preregistered following ethics approval and before participant recruitment. Further studies will subsequently evaluate the platform in conjunction with standardized olfactory and cognitive measures, subject to appropriate ethics approval. Conclusion: Sniff It provides a digital framework connecting structured olfactory training with personal scent tracking and everyday smell journaling. Its intended contribution is not to replace standardized clinical smell testing, but to make repeated home training and longitudinal self-observation easier to perform and document. Empirical evaluation is required to determine its effects on adherence, usability, and clinical outcomes. About this preprint: This system-design paper documents the rationale, architecture, implementation, and design boundaries of Sniff It Version 1, including the system demonstrated publicly in September 2026. It does not report participant data or clinical-effectiveness results. Prospective evaluation of usability, acceptability, and self-directed use is planned. Supplementary materials: Screenshots and demonstration videos for Sniff It Version 1 are archived separately at DOI: 10.5281/zenodo.22825622.