Skip to content

spatial-computing.md

Latest commit

 

History

History
112 lines (84 loc) · 3.42 KB

spatial-computing.md

File metadata and controls

112 lines (84 loc) · 3.42 KB
Raw

Spatial Computing

Spatial computing in Spanda covers shared workspaces, spatial anchors, and overlay workflows that connect humans, wearables, AR/VR/XR devices, and robots — implemented through optional packages and solution blueprints, not core rendering engines.

Related: ar-vr-xr.md · hri.md · solutions/spatial-computing.md

Concepts

Concept Description Implementation
Spatial anchor Fixed point in physical or digital space Package provider; synced via MQTT/ROS2/WebSocket
Shared workspace Multi-user spatial session Control Center session registry + package bridge
Robot overlay Live robot pose, path, sensor frustum in AR spanda-arkit / spanda-hololens + twin mirror
Mission overlay Waypoints, tasks, approval state in AR Mission continuity + readiness state export
Sensor overlay Lidar/camera frustums, hazard zones Device tree + assurance anomaly hints
Readiness overlay Go/no-go indicators per asset Readiness API → AR provider
Digital twin overlay Twin state rendered in XR Existing twin mirror + replay buffer

Spatial session lifecycle

1. Operator readiness passes
2. AR device registers session (package provider)
3. Spatial anchors published to shared workspace
4. Robot / mission state streamed to overlay
5. Context events (geofence, hazard) update overlay + robot
6. Session ends → audit + replay capture

Sessions are not managed by the core runtime — packages implement SpatialSessionProvider (planned provider trait) and register with the provider dispatch table.

Shared workspaces

A shared workspace binds:

  • One or more humans (operators, remote experts)
  • AR/VR/XR devices per participant
  • Robots / drones under collaborative mission
  • Control Center as session authority
[[spatial_sessions]]
id = "maint-session-001"
workspace = "bay-3-repair"
participants = [
  { human_id = "tech-001", device_id = "hololens-001", role = "field" },
  { human_id = "expert-002", device_id = "desktop-cc", role = "remote_expert" },
]
robot_id = "arm-001"
mission_id = "guided-repair-17"

Context awareness

Spatial computing composes with fleet geofencing and readiness alerts:

Operator enters hazard zone
  → Wearable alert (spanda-industrial-wearables)
  → AR warning overlay (spanda-hololens)
  → Robot slows (safety zone + mission update)
  → Mission state updated (continuity checkpoint)
  → Control Center incident (optional)

Supported context types:

  • Restricted areas and danger zones
  • Geofencing (GPS, UWB, beacon)
  • Proximity alerts (robot ↔ human)
  • Fatigue alerts (optional health)
  • Emergency evacuation workflows

Device tree

[[fleet.ar_devices]]
id = "hololens-001"
type = "ARHeadset"
provider = "spanda-hololens"
human_id = "tech-001"
capabilities = ["spatial_anchors", "hand_tracking", "robot_overlay", "annotation"]

[[fleet.vr_devices]]
id = "quest-training-001"
type = "VRHeadset"
provider = "spanda-openxr"
capabilities = ["vr_training", "mission_replay", "digital_twin_view"]

Control Center

  • AR Session Viewer — active sessions, participants, anchor list
  • Live Collaboration — participant graph, mission allocation

See control-center.md.