Getting Started with Aeion Spatial

Navigate to Spatial → Scenes and click New Scene.

Required fields:

• Name — e.g., "Summer Collection Showroom"
• Scene Type — Choose the best fit for your use case:

Scene metadata:

json { "name": "Summer Collection Showroom", "sceneType": "generic", "status": "draft", "settings": { "ambientLight": { "intensity": 0.4, "color": "#ffffff" }, "environment": "studio", "camera": { "fov": 50, "near": 0.1, "far": 1000 } } }

Once created, the scene opens in the Spatial Editor — a React Three Fiber canvas with:

• Left panel: Hierarchy tree of all scene nodes
• Center: 3D viewport with orbit/pan/zoom controls
• Right panel: Inspector for selected node properties (transform, material, physics)

Editor controls:

• Left-click: Select object
• Right-click: Context menu (duplicate, delete, group)
• Middle-click drag: Orbit camera
• Scroll: Zoom
• F key: Frame selected object
• Ctrl+Z: Undo

Instead of uploading, generate a model from a text description or reference image:

Open: Spatial → AI Generation

Text-to-3D

1. Select Text to 3D tab
2. Enter a prompt: "A photorealistic leather office chair with chrome base"
3. Select provider:
4. Select output format: glb, gltf, fbx, obj, usdz
5. Click Generate

Generation job tracked in your AI job history:

json { "id": "ai_job_01J8XK4N3P", "provider": "luma", "jobType": "text_to_3d", "status": "processing", "input": { "prompt": "leather office chair with chrome base", "resolution": "2048" }, "createdAt": 1746969600000, "completedAt": null }

Processing time: 10-120 seconds depending on provider and complexity. You'll receive a notification when complete.

Image-to-3D

Upload a reference photo → the AI reconstructs a 3D model from it. Best for:

• Physical product photos → 3D product viewer
• Architectural photos → 3D reconstruction
• Sculptures/artwork → 3D archive

Texture Generation

Generate PBR texture sets (diffuse, normal, roughness, metalness, AO) from text:

• Prompt: "weathered concrete with moss growing in the cracks"
• Output: 4K PNG textures or EXR HDR maps
• Used to create PBR materials automatically

Style Transfer

Apply artistic style to existing textures: "make this look like watercolor painting"

Publish the Scene

Click Publish in the scene editor toolbar.

What happens:

1. Scene nodes validated (no broken references)
2. All assets compressed and streamed to CDN
3. LOD variants generated for mobile delivery
4. spatial.scene.published event fires
5. Scene is publicly accessible (if permissions allow)

json { "id": "scene_01J8XK4N3P", "status": "published", "publishedAt": "2026-05-11T10:30:00Z", "embedUrl": "https://yourtenant.aeionos.com/spatial/scene/summer-collection", "viewerUrl": "https://yourtenant.aeionos.com/spatial/view/summer-collection" }

Create an Embed Config

Spatial → Embeds → New Embed

json { "sceneId": "scene_01J8XK4N3P", "allowedDomains": ["yourstore.com", "yoursite.com"], "showToolbar": true, "showAnnotations": false, "expiresAt": "2027-01-01T00:00:00Z", "maxViewCount": 10000, "encrypted": false }

Embed in a Commerce Page

html <!-- Auto-generated embed snippet --> <iframe src="https://yourtenant.aeionos.com/spatial/embed/e8f4a7b2..." width="100%" height="600" allow="xr-spatial-tracking; fullscreen" style="border: none; border-radius: 12px;" ></iframe>

Use case in Commerce: Product pages with an interactive 360° viewer. The viewer is linked to the product record — inventory levels, price, and variants all update live. No manual syncing.

For VR (Meta Quest, HTC Vive, etc.)

1. Open the published scene URL in a WebXR-compatible browser (Meta Browser on Quest, Firefox Reality, etc.)
2. Click the VR headset icon (bottom-right of viewport)
3. Browser prompts for WebXR permission → Allow
4. Scene renders in stereo VR — 6DOF head tracking, hand controllers mapped

VR controls:

• Left thumbstick: Teleport / smooth locomotion
• Right thumbstick: Turn
• Grip button: Grab and move objects
• Trigger: Select / interact
• Menu button: Open scene menu

For AR (Phone/Tablet)

1. Open the scene URL on a mobile device with AR capability
2. Click the AR icon (bottom-right)
3. Camera activates → point at surface → tap to place scene
4. Scene anchored to physical surface via ARKit (iOS) or ARCore (Android)

AR features:

• Surface detection (horizontal and vertical planes)
• Image tracking (trigger AR experience by pointing at a poster)
• Light estimation (scene lighting adapts to physical environment)

Generate a Launch QR Code

From the scene editor: Share → Generate QR Code

``` Deep link: aeion-spatial://scene/summer-collection?mode=vr QR code generated from that deep link.

// QR scanned on phone: // - iOS: Opens in Safari → AR Quick Look or WebXR // - Android: Opens in Chrome → WebXR AR mode // - Quest: Opens in Meta Browser → Full VR ```

Invite Collaborators

Share → Invite Collaborators

json { "sceneId": "scene_01J8XK4N3P", "invitees": ["user@yourtenant.com"], "permission": "editor" // "viewer" | "editor" | "admin" }

The invited user receives an email with a direct link. When they open the scene, they see you editing in real time.

What You See

• Other users' cursors — labeled with their name, showing position in the 3D view
• Selection highlights — when someone selects an object, you see a colored outline around it
• Camera views — see where other users are looking (optional, toggleable)
• Presence indicators — avatar for each user in the top toolbar

Real-Time CRDT Sync

Powered by Yjs CRDT — no central server authority:

• Every edit (transform change, material update, node add/delete) is a CRDT operation
• Operations merge automatically — two people editing the same material simultaneously? Both changes apply correctly
• Undo/redo replays across all participants — your undo undoes your last operation, not theirs
• Object locking — if you're editing a node's material, others see a lock icon and can't edit until you're done

Offline support: Disconnect and reconnect — Yjs replays your changes when back online.