Project: Phygital Wellness Kitchen
The Phygital Wellness Kitchen puts fine-dining Mediterranean nutrition into a single smart, climate-controlled unit that runs 24/7: the customer describes what they want in plain language, an AI engine builds a tray from what is in stock, and the meal is delivered through an ATM-style slot. Source: "Mediterranean Wellness Kitchen" (SFS paper, Oct 2026).
1. The "1+1+1 = 111" triad
| Pillar | Idea taken | Role in the unit |
|---|---|---|
| Pret A Manger efficiency | Compress four temperature cabinets (hot, cold, salad, snacks) into one multi-zone unit | The SFS paper estimates 80%+ footprint compression VMA assumption |
| DO & CO culinary standard | Central gourmet production and chef standards | Daily supply of meze, hummus, proteins, pita |
| OnQ Solutions hardware | Retail display, smart cabinetry, touchscreen UX | Chassis, climate dispensing, locking, interface |
VMA note. DO & CO is named in the SFS paper as a culinary model; there is no relationship with DO & CO and it has not been approached. In this plan the culinary pillar is supplied by the platform's own central kitchen (Azusa / Ohanyan's), with a catering partner as an alternative. OnQ's scale figures in the SFS paper come from that paper and have not been independently verified here.
2. Customer journey
- Ask: natural-language request at the touchscreen (example in the paper: "very hungry, under 500 kcal, pescatarian").
- Build: the AI engine matches stock, allergens and macros and shows the tray with calories and protein.
- Craft: in flagship store-in-store units a chef assembles the tray at an open marble counter (the Mediterranean Chef layer, Section 07).
- Deliver: contactless motorized slot; a "Tray Match" badge confirms the macros.
3. Target locations
| Vertical | Problem solved |
|---|---|
| Hotel lobbies | 24/7 room-service replacement without night-shift kitchen labor |
| Hospitals | Fresh, macro-verified food for night shifts, patients and visitors instead of vending |
| Universities & student housing | Off-hours meals; meal-plan card integration |
| Luxury residential | Food as a building amenity |
| Senior living | Diet-compliant (e.g., low-sodium Mediterranean) meals without kitchen overhead |
| Supermarket store-in-store | Replaces labor-heavy deli and hot bars; sits next to the Meze Bar |
4. Business model (Akdeniz Bar AI Smart Kitchen plan)
Source: Akdeniz Bar — AI Akıllı Mutfak & Sağlıklı Öğün İş Planı (TR, Oct 2026).
| Element | Plan |
|---|---|
| Placement (B2B2C) | Revenue share with the host (hotel, hospital, supermarket) or fixed rent in high-traffic sites |
| Back-end AI | Demand forecasting by site, weather, time and foot traffic; dynamic pricing near expiry; unsold food donated to local charities (ESG criterion for hosts) |
| Revenue | Fresh-food sales + data / software (SaaS) services |
| Target KPIs | Gross margin 60–65%; payback 12–18 months in high-yield sites; unit CAPEX to be optimized with OnQ VMA assumption |
| Roadmap | Phase 1 (months 1–6): prototypes, UI/UX, pilot sites (hotel, hospital, supermarket), AI calibration · Phase 2 (6–18): regional roll-out in the five verticals · Phase 3 (18+): scale and SaaS |
VMA note. The plan names Pret A Manger, DO & CO and The Pret Foundation as pillars; none has been approached and there is no relationship. Gross-margin and payback figures are plan targets, not yet tested.
5. What must be proven
- Unit economics: unit CAPEX, daily trays, waste rate — the 60–65% margin and 12–18-month payback targets must be tested with an OnQ quote and pilot data.
- Regulation: retail food and vending permits vary by state and county; to be confirmed with local health authorities before pilots.
- Pilot plan (VMA proposal): three sites in Southern California — one hospital, one hotel, one store-in-store — supplied from the central kitchen VMA assumption.