AvalaAvala
Start Building — Book a Demo / Become a Coworker

MCAP and ROS annotation

Annotate MCAP and ROS robot runs in one workflow.

Upload MCAP, ROS 1 bag, or ROS 2 bag recordings, inspect synchronized sensor streams, route selected scenes through labeling and QC, and export model-ready data through the API or SDK.

Start BuildingBook a Demo

The workflow

From robot recording to training release.

Avala keeps visualization, annotation, review, and export connected to the source dataset instead of splitting robot runs across disconnected tools.

  1. 01

    Ingest

    Bring in .mcap, ROS 1 .bag, and ROS 2 .db3 recordings, including camera, LiDAR, and telemetry messages.

  2. 02

    Visualize

    Scrub synchronized multi-sensor sessions in the browser before selecting the scenes that need labels.

  3. 03

    Annotate and review

    Apply 2D and 3D annotation workflows, route work through review, and keep QC state visible.

  4. 04

    Export

    Use the REST API or SDK to export annotations for your training pipeline.

Developer paths

Continue in the documentation.

Read the MCAP and ROS integration guide, review annotation tools, or call the API from your pipeline.

MCAP and ROS guideOpen docsAnnotation toolsOpen docsAPI referenceOpen docs