> ## Documentation Index
> Fetch the complete documentation index at: https://docs.cyberwave.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Teleoperation & Remote Control

> Stay in the loop with your physical AI — watch every twin live, and take over any robot in one click

Cyberwave is built for both research projects **and** the moment those projects ship to a real customer. Once a robot is doing real work in a real place, **human override is a requirement, not a feature** — and Cyberwave gives you that override out of the box for every twin in the [catalog](https://cyberwave.com/catalog), with no extra plumbing.

```python theme={null}
from cyberwave import Cyberwave

cw = Cyberwave()
cw.affect("live")

robot = cw.twin("acme/twins/arm-station-1")
robot.use_controller("keyboard")          # human takes the wheel
```

The robot was being driven by a model a second ago; now it's being driven by you. Same twin, same environment, same code path.

***

## See your fleet in real time

Switch any [environment](/overview/features/environment) into **Live** mode and the dashboard becomes a window into the physical cell. You don't tab between half a dozen camera apps — Cyberwave composes the picture for you:

* **Live video** from every paired camera, streamed via WebRTC into the same canvas as the 3D scene.
* **3D twin context** — robot poses, joint states, and waypoints animate live alongside the video so you always know what's where.
* **Spatial sensors** — [LiDAR and depth point clouds](/feature-reference/environment-editor/index#live-point-cloud-alignment) overlaid on the scene so you can see what the robot sees.
* **Health at a glance** — per-twin [presence indicators](/feature-reference/environment-editor/index#twin-presence-indicator-live-mode) and edge-core status make it obvious which robot needs attention.

See [Environment Editor → Live mode](/feature-reference/environment-editor/index#edge-core-connection-status-live-mode) for the full panel.

***

## Take over any robot in one click

Every twin always has exactly one **controller** driving it. Swapping controllers is the entire takeover flow — there's no separate teleoperation app to launch.

<CardGroup cols={2}>
  <Card title="Keyboard" icon="keyboard">
    Drive joints, gripper, or base motion straight from the browser. Best for
    quick recoveries and demos.
  </Card>

  <Card title="Gamepad" icon="gamepad">
    USB or Bluetooth controllers for intuitive control of mobile robots and
    arms.
  </Card>

  <Card title="Leader arm" icon="hand">
    Pair a physical leader arm (e.g. SO-101 leader/follower) and mirror your
    motion 1:1 onto the live robot.
  </Card>

  <Card title="Python SDK" icon="python">
    Script overrides from your own code with
    [`cw.affect("live")`](/tools/python-sdk#affect-simulation-vs-live-environment)
    and `robot.use_controller(...)`.
  </Card>
</CardGroup>

```python theme={null}
robot.use_controller("keyboard")          # human override
robot.use_controller("my-vla-policy")     # hand it back to the AI
```

Telemetry, recording, and the [environment replay](/feature-reference/environment-editor/replay) timeline keep running across the handoff — a single session can start with an AI policy, hand off to a human, and end with a code controller, all on the same twin.

Full reference: [Live Teleoperation](/feature-reference/environment-editor/teleoperation).

***

## Alerts, escalation, and who can do what

Real production deployments need more than "someone is watching the dashboard". Cyberwave wires the rest of the loop for you:

<CardGroup cols={2}>
  <Card title="Real-time alerts" icon="bell" href="/feature-reference/edge/drivers/alerts">
    Per-twin and per-environment alerts with `info`/`warning`/`error`/`critical`
    severities, acknowledgement, and resolution lifecycle — surfaced in the UI
    and over [MQTT](/api-reference/mqtt/main).
  </Card>

  <Card title="Workflow-driven escalation" icon="diagram-project" href="/overview/features/workflows">
    Use [workflows](/overview/features/workflows) to fan out alerts to email,
    [HTTP](/feature-reference/workflows/overview#nodes), or your on-call system
    when a model gets stuck, a confidence drops, or a sensor goes silent.
  </Card>

  <Card title="Role-based access" icon="shield" href="/feature-reference/access-control">
    [Viewer / Editor / Admin](/feature-reference/access-control) roles
    cascade from the workspace, with per-twin and per-environment overrides
    so operators only see (and drive) what they're meant to.
  </Card>

  <Card title="Audit-friendly link shares" icon="link" href="/feature-reference/access-control">
    Read-only and comment-only [link shares](/feature-reference/access-control)
    let stakeholders watch a live cell without ever getting takeover rights.
  </Card>
</CardGroup>

Together, this is how you take a physical AI automation from "cool demo" to "I can put it in front of a paying customer" — full real-time visibility, instant override, and a defined chain of who responds when something is off.

***

## What to read next

<CardGroup cols={3}>
  <Card title="Live Teleoperation reference" icon="gamepad" href="/feature-reference/environment-editor/teleoperation">
    Controller types, WebRTC streaming, latency, and the full setup flow.
  </Card>

  <Card title="Go2 — digital to physical" icon="dog" href="/tutorials/go2-digital-to-physical">
    End-to-end walkthrough of swapping a simulated policy for live override.
  </Card>

  <Card title="SO-101 teleop dataset" icon="record-vinyl" href="/tutorials/so101-teleop-dataset">
    Use a leader/follower setup to teleop a real arm and record data for
    training.
  </Card>
</CardGroup>