ROS 2 Editor

The ROS 2 Editor is a World Editor tool for building per-vehicle ROS 2 sensor configurations inside BeamNG.tech, launching the in-process low-latency native C++ ROS 2 bridge for those vehicles, and saving/loading reusable JSON templates. It complements the external BeamNG–ROS 2 packages by letting you author and run a vehicle’s published sensor set from the simulation UI.

The bridge runs inside the simulator process: sensor data is published directly from the engine through the loaded ROS 2 shared library, without a Python relay or socket round-trip. That avoids the extra copy/serialisation steps of the external Python integration, so topics stay low-latency and can be published at high rates (subject to each sensor’s update limiter).

ROS 2 Editor with library loaded and sensor type grid

Prerequisites

The remote beamng_ros2 Python packages remain available for scenario orchestration; this editor focuses on in-simulator library loading, sensor draft editing, and per-vehicle start/stop.

Opening the editor

1. Open the World Editor.
2. In the mode bar, select ROS 2 Editor (device-hub icon).
3. The ROS 2 Editor window opens.

ROS 2 library

At the top of the window:

• The current library path is shown (green when loaded).
• If nothing is loaded, use Browse… to pick the native C++ bridge library, or Load last used to reload the path stored in editor preferences (ros2Editor.general.lastRosLibPath).

Until the library loads successfully, the rest of the editor body is disabled. An incompatible library version surfaces an error in this section.

Scene vehicles

The Scene vehicles table lists every active vehicle. Green row tint means ROS is currently running for that vehicle.

Per-row actions:

ROS is started per vehicle. You can run several vehicles independently, each with its own sensor draft.

Adding and managing sensors

Below the vehicle table:

Supported types in the defaults catalogue:

Exact availability follows tech/rosDefaults.json shipped with the build (copied to the user tech/ folder on first use when needed).

Sensor inspector

With a sensor selected, the inspector edits the draft before launch.

ROS 2 topic / identity

Scalar channels

For scalar sensor types (Electrics, State, and similar):

Placement (positional sensors)

Sensors with pos / dir / up in their create config show a placement section:

Rates and publishing

Blob / GPU-backed types expose:

Sensor parameters

Type-specific controls (examples):

Parameters not given a dedicated control still appear in the generic parameter list when present on the sensor’s default create table.

JSON templates

Save template writes a launch-ready JSON document:

{
  "sensors": [
    {
      "id": "Camera_1",
      "type": "Camera",
      "config": { "...": "..." },
      "fields": [ "optional subset of scalar field names" ]
    }
  ]
}

The export merges each draft sensor with its type defaults (including message metadata from rosDefaults.json). If scalar subset mode is active, only the selected fields are kept.

Load template replaces the selected vehicle’s draft with the file contents. Sensors that already list fields are treated as subset mode on load.

Typical workflow

1. Load the ROS 2 library once per editor session (or use Load last used).
2. Spawn the ego (and any other) vehicles.
3. Select a vehicle in the ROS 2 Editor.
4. Add sensor types, set ids, placement, rates, and optional scalar subsets.
5. Optionally save a JSON template for CI / reuse.
6. Press Launch on the vehicle row.
7. Consume topics from your ROS 2 graph; press Stop when finished, or Reload after editing the draft.

Related documentation

• ROS 2 interface overview — external beamng-ros2-integration packages and messages.
• ADAS Sensor Configuration Editor — vehicle-attached sensors for BeamNGpy / co-simulation / VSL (separate from the ROS draft, though many sensor concepts overlap).
• Sensors — behaviour and parameters of individual BeamNG.tech sensors.