Rerun & MCAP
OmniLoop positions itself as the control plane to observability tools like Rerun and Foxglove. Rather than competing on rich 3D visualization, OmniLoop makes them complementary.
Rerun Integration
Section titled “Rerun Integration”The RerunSink forwards every OmniLoop telemetry frame directly into Rerun as time-series scalars.
from omniloop.integrations.rerun import RerunSink
loop.add_sink(RerunSink(app_id="star1"))Installation
Section titled “Installation”pip install omniloop[rerun]MCAP Conversion
Section titled “MCAP Conversion”If you have historical robot data stored in MCAP logs, you can import them into an OmniLoop .omni journal for playback and analysis using mcap_to_journal.
from omniloop.integrations.rerun import mcap_to_journal
frames = mcap_to_journal("recording.mcap", "imported.omni", topics=["/imu"])Installation
Section titled “Installation”pip install omniloop[mcap]Conversion Features
Section titled “Conversion Features”- Transparent JSON and ROS 2 CDR decoding.
- Optional topic filtering to only import the data you care about.
- Parameters:
app_id: Application identifier.spawn: Whether to spawn a viewer immediately.prefix: Prefix added to topic names.skip: List of fields or topics to exclude from the import.