○ Olfactory-Led Personalized Multisensory Reminiscence Content and Multibiomarker-Based XAI for Cognitive Health Management, OSMIA - Olfactory-Led Personalized Content: RFID retrieves user history and delivers a personalized one-minute reminiscence experience with visual, auditory and synchronized olfactory cues, encouraging spontaneous verbal recall - AI biomarker inputs: multi-wavelength fNIRS captures task-related HbO/HbR changes; speech analysis derives SPM, pitch, jitter, shimmer, formant, and intensity - Multibiomarker-based XAI: XGBoost integrates physiological and behavioral biomarkers, presenting cognitive-health indicators and session-level feature contributions alongside personalized guidance and longitudinal trends - Fast, private operation: internal benchmark processing a 14.6-second speech sample in 0.7 seconds; raw speech kept local, deleted after one day - Field-tested usability and engineering: 391 adults aged 50+; 4.2/5 average usability; 48.4 dB(A) operating noise; MFDS Class I status limited to the integrated fNIRS probe - Protected differentiation: Four Korean patents and a PCT application related to digital scent and olfactory cognitive-health technologies