← Innovation Labs/Lab L3ResearchImmersive & spatial

Spatial Commerce Lab — Vision Pro · Quest · WebXR · NeRF · Gaussian Splatting

Spatial isn't a gimmick — it's the next interaction layer for commerce, training and field ops. The Lab ships production-grade pipelines for capture, reconstruction, governance and distribution across Vision Pro, Quest and WebXR, with measurable conversion and skill-transfer outcomes.

Research thesis

When the buying or training context maps 1:1 to the physical world, conversion and skill transfer step-change.

Conversion uplift

+ 31%

Time-to-competency (training)

− 47%

Capture-to-publish

21d → 36h

Become a design partner →See live experiments

Active experiments

What the lab is testing right now

Gaussian Splatting at retail scale

Capture pipelines for entire stores; sub-second loads on Quest 3 and Vision Pro.

Try-on with consented likeness

Per-user avatars with consent ledger; never trained on, never persisted beyond session.

Field-ops digital twins

1:1 plant twins for safety, training and remote expert assist.

Spatial telemetry

Gaze, dwell, hand-pose analytics — anonymised, privacy-preserving, measurable.

Shippable artefacts

Everything the lab ships

Capture rig + pipeline
Reference rig, photogrammetry pipeline, NeRF / Splat training, asset CDN.
WebXR runtime
Cross-device experience layer for Vision Pro, Quest, Android XR and browser fallback.
Spatial CMS
Author, version and govern spatial scenes; A/B test like web pages.
Consent + telemetry SDK
Per-session consent capture and privacy-preserving spatial analytics.
Reference experiences
Storefront, training bay, plant twin and museum patterns ready to fork.

Lab team

Spatial Principal
Computer Vision Engineer (NeRF / Splat)
WebXR / RealityKit Engineer
UX / Interaction Designer

Partners we collaborate with

Apple Vision ProMeta Reality LabsNiantic 8th WallUnityUnrealWebXR W3C

Example output · Scene · storefront.lhr.flagshipjson

{
  "scene":      "storefront.lhr.flagship",
  "version":    "2026.05.01",
  "capture":    { "method": "gaussian_splatting", "frames": 1842 },
  "size_mb":    18.4,
  "devices":    ["visionpro", "quest3", "androidxr", "webxr"],
  "interactions": [
    { "id": "tryon.shoe.zr-90", "consent_required": true },
    { "id": "ar.staff.assist",  "remote_expert": true   }
  ],
  "telemetry": {
    "gaze_heatmap": true,
    "dwell_zones": ["shelf-A", "shelf-B", "checkout"],
    "anonymise":   "k=8"
  }
}

Engagement timeline

Weeks 1–10 · first WebXR pilot live by week 4

1
Weeks 1–3
Capture + reconstruct
On-site capture, photogrammetry / Splat pipeline, asset CDN ready.
2
Weeks 3–6
Cross-device runtime
Vision Pro + Quest + WebXR experience live with consent + telemetry.
3
Weeks 6–10
Measure + iterate
A/B uplift on conversion or competency, spatial CMS workflow embedded.

Flagship pods

Productionised by these squads

Spatial Storefront Pod

Field Training Pod

Plant Twin Pod

Museum & Heritage Pod

Selected publications

Receipts, not just thesis

Gaussian Splatting beats NeRF on retail-scale capture pipelines
SIGGRAPH Asia (poster)·2025
Conversion uplift in spatial commerce: a 14-brand study
AXP Internal Whitepaper·2026

FAQs

What partners actually ask

Do we need a Vision Pro budget per user?

No — WebXR fallback covers 95% of audience on existing devices; Vision Pro and Quest are premium tiers.

How do you handle likeness consent?

Per-session consent ledger, ephemeral avatars, no model training on user likeness, audit log on every interaction.

File sizes?

Gaussian Splatting + streaming LODs hit sub-20MB scenes for retail-scale captures.

Is this AR or VR?

Both — and increasingly the same runtime. We target a single experience that adapts to device and context.

Design-partner programme · L3 Spatial Commerce Lab

Co-build Spatial Commerce Lab with us in Weeks 1–10.

We'll respond within one business day with a scoping note, a fixed-price outcome contract, and a named principal cleared for your domain. Design partners get first-look access, joint publication rights and roadmap influence.

• Outcome-priced — no T&M.
• Sovereign by default — your data, your region, your keys.
• Refund-backed if the contracted KPI isn't hit.
• Joint publication rights and conference slots.