BOAXR · DEEPTECH · 01

Spatial intelligence,
for the real world.

FritzNoMo is BOAXR's enterprise spatial computing platform — a deeptech engine that turns physical environments into addressable, machine-readable space. Built for the operators, engineers, and frontline teams who need mixed reality to actually work in production.

Status// In Stealth
Release// Limited Pilots, 2026
Sector// Industrial XR
Hardware// Headset-agnostic

Capabilities

Six things FritzNoMo does that nothing else does together.

A unified deeptech runtime — perception, persistence, collaboration, and edge inference — designed to be deployed once and survive whatever your facility throws at it.

// PERCEPTION

Persistent 3D Scene Graph

Reconstructs and remembers your environment across sessions, devices, and lighting conditions — no markers, no re-scan rituals.

// ANCHORING

Sub-centimeter World Lock

Holograms stay where you put them. Across days, weeks, restarts, and different headsets walking into the same room.

// COLLABORATION

Co-presence at Scale

Dozens of users in the same physical room — or remote — viewing and manipulating the same shared scene in real time.

// INFERENCE

On-device Vision Models

Object recognition, defect detection, and pose estimation running at the edge — no round trips to a server, no data leakage.

// INTEROP

Headset-agnostic SDK

One codebase, every device. Quest, Vision Pro, HoloLens, Pico — and the next generation, whatever that turns out to be.

// SECURITY

Air-gapped Deployments

Run the entire stack inside your perimeter. Spatial data, telemetry, and inference never leave your network.


Architecture

Four layers. One runtime.

FritzNoMo is built as a clean stack — sensing at the bottom, applications at the top — so teams can plug in at the layer that fits their use case.

// LAYER 04
Experience SDK
Unity, Unreal, WebXR bindings. Build apps in days, not quarters.
Stable
// LAYER 03
Collaboration Mesh
Low-latency state sync across clients. Self-healing topology.
Beta
// LAYER 02
Spatial Memory
Persistent scene graph + relocalization. The part that makes things actually stay put.
Stable
// LAYER 01
Perception Core
Sensor fusion, SLAM, on-device inference. Hardware-agnostic.
Proprietary

Each layer is independently addressable. Use the full SDK to ship a turnkey experience, or drop down into the perception core if you're building something we haven't imagined yet.

Everything compiles to a single runtime that ships to the headset. No bloat, no surprise dependencies, no internet required at runtime.

// quick start — anchor a hologram to real space
import { Scene, Anchor } from '@fritznomo/sdk';

const scene = await Scene.load('factory-floor-3a');
const anchor = scene.anchor(myHologram, { persistent: true });

// it'll be there tomorrow. and the day after that.

By the Numbers

Engineered to production tolerances.

The benchmarks that matter when MR moves from demo to deployment.

<6mm
Anchor Drift
Over 24 hours, across re-localizations.
90fps
Render Floor
Sustained on mobile-class chipsets.
32
Co-presence
Concurrent clients per scene mesh.
0ms
Cloud RTT
Fully on-prem. Optional cloud sync.

Where It Ships

Built for the places where reality is non-negotiable.

FritzNoMo isn't built for the metaverse. It's built for floor plans, work cells, training bays, and the people who maintain them.

// MANUFACTURING

Guided Assembly & QA

Overlay step-by-step instructions onto the part in front of you. Catch defects the moment they happen, not after the line moves.

→ Pilots in progress
// FIELD SERVICE

Remote Expert Co-presence

An engineer in Bengaluru can stand inside the machine a technician in Hamburg is repairing — and point at the bolt that matters.

→ Pilots in progress
// TRAINING

High-stakes Skill Transfer

Repeatable, instrumented training scenarios that run on real equipment without the risk, the downtime, or the consumables.

→ Design partners welcome
// FACILITIES

Digital Twin, Walkable

Bring the BIM model, the IoT telemetry, and the floor itself into a single shared layer that anyone with a headset can step inside.

→ Roadmap, 2026

Currently In Stealth · Limited Cohort

Tell us what you'd build with it.

We're picking a small group of design partners for the first wave of pilots. If you have a problem that needs spatial computing to actually work — and you're willing to break it with us — we want to hear from you.