VeluneOS Compare
Compare a Reference run and a Target run of your robot. Get a stated decision and the structural differences to review, each with its time window.
- Beta users
- Sign in to Compare with your Beta credentials → upload
comparison.velune - New to Compare
- Request access → approval and written data-use terms → credentials
- What
- Recorded message interval longer in Target — Stream 003
- When
- T+10s → T+20s
- Evidence
- median interval 100.0 → 200.0 ms · returned to the Reference pattern at T+20s · Comparable
Use it
Prepare locally. Upload one file.
- Install the Adapter Kit on the machine with your recordings (Python 3.12 or 3.13).
velune mapping initon the Reference and Target MCAP; label streams and confirm any pairing the Adapter asks about.velune preparewritescomparison.veluneand a preview of everything inside it.- Sign in to Compare, upload
comparison.velune, download the report.
Raw MCAP is not uploaded; Compare accepts only comparison.velune. Velune Trace is not required. Step-by-step guide
What the report contains
- DecisionREVIEW_REQUIRED, NO_MATERIAL_STRUCTURAL_CHANGE, INCONCLUSIVE_DIFFERENCE or INSUFFICIENT_COVERAGE
- FindingsStreams in one run only, interval changes, activity loss and recovery, timestamp representation changes
- For each findingIts window, values from both runs and evidence quality
- OutputHTML, Markdown and JSON, downloaded in the browser
A review order, not a severity score. Compare does not decide root cause, fault, safety, regression or release readiness. Reading the report
Request Beta access
Start from a real change.
Tell us what changed between two runs and what you need to know. During the Private Beta we work through comparisons with each team directly.
Access options: Contribution (no fee — approved derived data may be used to develop and improve VeluneOS products and services) or Private (quoted per team — your data is used only to provide the service you requested). Choosing an option here is not consent.
Already in the Beta? Sign in to Compare.