AR in Food and Beverage: Digital Menus, Kitchen Training, and Spatial Restaurant Design

TL;DR:

• AR-enhanced menus on smartphones and smart glasses increase average order value by 20–40% in documented deployments — customers order more when they can visualise dishes
• VR kitchen training cuts onboarding time by 30–50% for complex cooking and food safety procedures, and is gaining traction with QSR chains and hotel groups
• Spatial digital twins are being used by restaurant groups to plan new outlet layouts and test kitchen configurations before construction begins

The food and beverage industry is rarely associated with cutting-edge technology adoption. But in 2025 and 2026, spatial computing applications — AR menus, VR staff training, and mixed reality kitchen planning — are moving from pilot programmes into regular deployment, driven by practical outcomes that justify the investment.

The use cases aren’t about gimmicks. They address real problems: reducing food waste from customer disappointment, cutting training costs in a high-turnover industry, and compressing the design iteration cycle for new restaurants.

AR Menus: More Than a Novelty

The concept of an AR menu — pointing your phone at a table marker and seeing a 3D model of a dish appear — has existed for years without catching on. Early versions were clunky, required app downloads, and offered low-quality models that did little to convey what a dish actually looked like.

What’s changed in 2025-2026 is a combination of factors: iOS and Android now render high-quality WebXR AR without an app install, 3D food visualisation tools have improved substantially, and cloud-based platforms (Mojo Vision, Aria, Avataar, and several smaller players) make it realistic for mid-sized restaurant groups to deploy AR menus without custom development.

The business case is increasingly well documented. A Hong Kong restaurant group running a three-month AR menu pilot across 12 outlets reported a 23% increase in average order value during the pilot period. A UK fine dining group using AR to showcase tasting menu courses reported a reduction in dish returns (customers sending back food that wasn’t what they expected) from 4% to under 1%.

The mechanism is simple: when customers can see what they’re ordering in realistic detail — the size, the presentation, the garnishes — they order more confidently and experience fewer disappointments. This is similar to the established e-commerce finding that product AR try-on reduces returns.

Current platforms work in two ways. Smartphone AR uses the phone camera and the platform’s web app or SDK to overlay 3D dishes on the physical table. No app download needed — the customer scans a QR code on the menu card and the AR launches in their mobile browser. Smart glasses AR is further ahead in enterprise settings: waitstaff using AR glasses can see customer preferences, allergen alerts, and dish specifications overlaid on their field of view while tableside.

For independent restaurants, the realistic entry point is smartphone AR. Platforms like Aria or Yum Brands’ internal AR system (licensed to third parties) offer subscription pricing that works at the single-outlet level. Expect to pay £200–500/month for a platform subscription and £100–300 per dish for professional 3D food scanning.

VR Kitchen Training

Staff turnover in food service is among the highest of any industry sector. In the UK, the Caterer estimates average hospitality turnover at 70–80% annually in fast casual and quick service. That means constant onboarding — and the standard onboarding model (shadow an experienced colleague, read a procedure manual, try things on live service) is slow, inconsistent, and expensive.

VR training offers a different model: standardised, repeatable training scenarios that a new hire can run through before touching live equipment or serving real customers.

McDonald’s, Compass Group, and SSP (Select Service Partner, operators of airport and travel hub food outlets) are among the larger UK operators who have run structured VR training pilots. SSP’s 2024-2025 programme, covering food safety and customer service scenarios, reported a 34% reduction in onboarding time for kitchen staff.

The training scenarios most suited to VR are those that involve procedural sequences with safety implications: food temperature checking, allergen handling procedures (especially NATASHA’s Law compliance in the UK), fire safety, and equipment operation for complex cooking processes like sous vide, combi oven programming, or coffee machine maintenance.

These scenarios work in VR because they require the trainee to follow a precise sequence, and getting it wrong in training has no consequences. On live service, a missed allergen check or an incorrect temperature recording has real consequences. VR creates a consequence-free environment to build the muscle memory and procedure recall before it matters.

Hardware for restaurant VR training is typically Meta Quest 3 headsets managed through Business Suite MDM, or purpose-built training platforms that include both hardware and content. Content development is the larger cost — building a realistic virtual kitchen environment with interactive equipment runs £15,000–50,000 for a full scenario set. That cost makes VR training economically attractive mainly for multi-outlet operators who can amortise it across many locations.

Spatial Digital Twins for Restaurant Design

A spatial digital twin is a 3D model of a physical space that can be experienced in VR or AR — walked through at 1:1 scale before the space is built. For restaurant groups planning new outlets or refurbishments, digital twins are replacing or supplementing traditional 2D floor plans and physical mockups.

The practical application: a restaurant group planning a 40-cover brasserie can build a spatial digital twin of the proposed layout and walk through it in VR to evaluate sightlines, server traffic patterns, kitchen-to-floor access, and ambiance — before any construction starts. Discovering that the kitchen pass is positioned in a way that creates server congestion during peak service costs nothing to fix in VR. The same discovery after fit-out costs tens of thousands of pounds.

Companies like Sketchfab, NVIDIA Omniverse (for large-scale projects), and purpose-built hospitality design tools from players like SiteMaster and CAD Chef are being used by restaurant architects and operators for this purpose. The technology for the underlying scan-to-model pipeline has also improved dramatically — a handheld LiDAR scan of an existing site (using an iPhone Pro or a dedicated scanner like the Matterport Pro3) can produce a usable 3D model in hours rather than weeks.

For chain restaurants standardising their outlet design, digital twins serve a second function: the master design twin is used to assess whether candidate sites (properties for potential lease) meet the design brief before spending on detailed surveys. A 3D scan of a prospective site can be overlaid against the standard design template to identify structural conflicts early in the site selection process.

What’s Practical for Independent Operators

The smart glasses and digital twin applications are primarily enterprise territory — viable for groups with 10+ outlets that can justify the development and management costs. For the independent restaurant owner, the accessible entry point is smartphone AR menus.

The barrier is lower than it was two years ago. Several platforms now offer no-code setup where you photograph your dishes with a guided scanning app and receive deployable 3D models within a few days. QR codes on physical menu cards link to the AR experience. No custom software development required.

For training, VR is currently most practical for businesses with multiple sites. A single-site independent restaurant owner is better served by digital training platforms (videos, interactive checklists) than VR investment. The economics shift as you open your second and third locations.

The underlying direction is clear: spatial computing tools are finding genuine purchase in food and beverage because the industry problems they solve — communication gap between what customers expect and what they receive, high-turnover staff training, expensive physical fit-out decisions — are large enough that even modest technology improvements justify the cost.