# vaos mobile

Vision-native runtime. Anywhere.

Monitor, test, stream, validate, and interact with VAOS cognition runtimes directly from mobile devices. Not a camera app — a runtime surface.

Developer Preview→ Key features

# what it is

The mobile runtime surface

VAOS Mobile extends the runtime ecosystem to phones and tablets for runtime monitoring, Scene Contract inspection, edge deployment, live cognition streaming, validation tracing, RTSP / WebRTC feeds, and portable agent control.

Camera / RTSP
        ↓
   VAOS Edge Runtime
        ↓
Scene Contract + Cognition
        ↓
   VAOS Mobile Runtime
        ↓
Dashboard / Alerts / Actions

The mobile app is part of the runtime network — not a standalone tool.

# key features

Runtime, not consumer UI

🗺️

Live Scene Contracts

Inspect structured perception output in real time, frame by frame.

{
  "risk": "medium",
  "motion": "approaching",
  "objects": ["person"]
}

📋

Runtime trace viewer

Watch Observe → Structure → Think → Validate → Act stream live. One of the strongest differentiators.

🔗

Edge runtime pairing

Connect to RK3588, Jetson, Pi 5, VAOS USB Runtime, robotics nodes. Plug, pair, monitor.

📡

RTSP / live stream

Drones, robots, smart cameras, mobile devices — all become runtime nodes.

🛡️

Validation alerts

"Validation blocked action: low sensor reliability." Trace why every decision passed or stopped.

🎯

Portable agent control

Pause, resume, escalate, or override agents from the field. Audit-logged.

# ecosystem integration

Connected runtime surfaces

⬡

VAOS Core Runtime

Native socket / WebSocket pairing.

🧳

Portable Runtime

USB-C device discovery and pairing.

🛠️

Edge Toolkit

RK3588, Jetson, Pi node monitoring.

🧠

SceneLM Runtime

Inspect reasoning traces live.

📨

MQTT & WebSocket APIs

Subscribe to scenes, actions, alerts.

🤖

Robotics runtime

Pause, escalate, override field agents.

# future

Runtime features in the pipeline

📷

Multi-camera fusion

Synchronised view across stereo + thermal + depth.

🎨

Live cognition overlays

AR-style reasoning visualisations on live feeds.

🌐

Distributed edge nodes

Aggregate cognition from multiple runtime nodes.

🎯

Predictive cognition

Forward-state visualisation from SceneLM-predictive.

🎤

Voice interaction

Talk to the runtime; receive structured responses.

🧪

SensorTwin sync

Simulate scenes locally for offline development.