Product · Robot Ops

Replay, incident detection, and failure triage for robot fleets.

Every robot failure is a training signal. Robot Ops turns raw robot logs into replayable incidents, AI-assisted triage, and a labeled ML dataset — the data flywheel that turns your robot's failures into your model's training advantage. Built on MCAP and Foxglove, deployed with a US robotics operator, with a runnable end-to-end demo.

logs → data
The flywheel
Raw MCAP logs become a labeled, queryable failure dataset your ML team can train on.
mcap-native
Foxglove replay
Records to MCAP — the ROS 2 native format Foxglove is built on — for frame-accurate incident replay.
triage
Evidence, not verdicts
Claude produces a triage hypothesis and investigation steps per incident — cached and reviewable, not a black box.
→ The pipeline

Raw logs in. A labeled dataset out.

Modern robotics teams have the logs — they don't have the pipeline to find the failures, replay them, explain what happened, and get structured labels into the hands of the ML team. Robot Ops is that pipeline.

It detects the incident window, replays the clean-vs-corrupted signal diff in Foxglove, generates a triage hypothesis with investigation steps, and exports a curated, annotated failure dataset — end to end.

// pipeline
RAW LOGS
mcap · ros2_bag · /scan · /scan_raw
INCIDENT DETECTION
detect_window · multi_signal · lidar_fault
REPLAY — FOXGLOVE
mcap_replay · clean_vs_corrupted_diff
AI TRIAGE (CLAUDE)
triage_hypothesis · investigation_steps
CURATION & HANDOFF
fiftyone_ingest · cvat_export
[ 01 ] Incident detection

Find the failure, not the haystack.

Detects the incident window across signals — a lidar fault, a navigation deviation, a perception dropout — so nobody scrubs hours of logs by hand looking for the moment it went wrong.

[ 02 ] Foxglove replay

Replay the exact moment.

Records both clean and corrupted signals to MCAP so your team can replay the incident in Foxglove — clean vs. corrupted, side by side — where the robotics ecosystem already converges for debugging.

[ 03 ] AI-assisted triage

A hypothesis, not a black box.

Claude produces a triage hypothesis and investigation steps for each incident, tied to the evidence. Output is cached and human-reviewed — never a live call in the loop, and never a "root cause" verdict your engineers can't check.

[ 04 ] Dataset & handoff

Failures become training data.

Frames around each incident are ingested into FiftyOne and exported as CVAT annotation tasks — a labeled, queryable failure dataset ready for your ML team, not a folder of unsorted images.

// see it run

Robot Ops ships with a runnable, end-to-end reference demo: a TurtleBot3 runs a warehouse mission in ROS 2 and Gazebo, a deterministic lidar fault is injected mid-run, the whole run records to MCAP, and the pipeline detects the incident, replays it in Foxglove, generates the triage hypothesis, and exports the annotated dataset — the full loop, on a dashboard you can open.

MCAP ROS 2 Humble Foxglove Nav2 · AMCL Gazebo Claude FiftyOne CVAT FastAPI
// where it stands

Robot Ops is deployed with a US robotics operator under NDA, and ships with a runnable reference demo of the full pipeline. It's US-focused today. We stand it up against your fleet's own data, document it, and hand it over — and can operate it for you if you'd rather we run it.

[ FAQ ]

Common questions about Robot Ops.

What robot log formats does Robot Ops work with?
The primary format is MCAP — the ROS 2 native format that Foxglove is built on. Robot Ops also ingests ROS 1 bags and structured telemetry exports from proprietary platforms. If a fleet generates data in a format we haven't seen, we scope a short ingestion sprint before the main engagement.
Does the triage step call an AI API live?
No. Robot Ops generates the triage hypothesis and investigation steps as cached, human-reviewable output — the dashboard serves cached results and does not call an AI API in the loop. The language is deliberately "triage hypothesis" and "investigation steps," not "root cause" or "diagnosis": the system surfaces evidence for your engineers to investigate, not a verdict they can't check.
Is Robot Ops a product or a service?
Both. Robot Ops is a productized pipeline — currently deployed with a US robotics operator under NDA, with a runnable end-to-end reference demo. We stand it up against your fleet's own data, hand it over documented, and can operate it for you if you want.
How does Robot Ops relate to the Physical AI Data Flywheel?
Robot Ops is the product; the Physical AI Data Flywheel is the engagement that stands it up — typically 90 days from raw logs to a queryable failure dataset. Same pipeline, same tools: MCAP, Foxglove, Claude, FiftyOne, CVAT.

→ Related reading

Physical AI Data Flywheel → Our robotics work → Edge & Perception AI → Production computer vision → EdgeTelemetry →

Point Robot Ops at your fleet's logs.

Robot Ops is deployed with a US robotics operator today, and we're taking on a small number of additional teams. If your robots are generating logs you're not learning from, we'd like to talk.