Fleet operations or release evaluation?
Use Formant or InOrbit when your immediate job is monitoring deployed robots, investigating incidents, and coordinating operational responses. Use OORB RunFP's recorded-episode workflow when you need baseline comparisons and explicit release checks; fleet analytics can also contribute evidence to that decision.
This comparison describes workflows, rather than declaring that fleet platforms cannot support evaluation. Vendor sources were checked on October 15, 2026.
What Formant and InOrbit document
Formant documents fleet observation, event-triggered actions, operator intervention, incident analysis, and data export in its incident-management guide. Its analytics documentation covers streams, events, and task summaries for historical analysis.
InOrbit's product documentation covers dashboards, robot coordination, remote interventions, and configuration updates. Its developer documentation describes Time Capsule for reviewing past incidents and mission performance.
These are useful sources of operational evidence. A fleet dashboard may help you identify a failure cohort or define a release KPI; the next step is connecting that evidence to a specific candidate, task, and acceptance policy.
Compare the decision you need to make
| Decision | Fleet operations workflow | RunFP's recorded-run workflow |
|---|---|---|
| Which deployed robots need attention now? | Inspect fleet status, alerts, and incidents in Formant or InOrbit | Analyze supplied run artifacts when investigating a particular behavior |
| Which field conditions coincide with failures? | Review available incident history, telemetry, and task or mission analysis | Compare healthy and failed episodes while recording task conditions alongside the experiment |
| Did this candidate differ from our reference run? | Use available analysis or connect an evaluation workflow to the platform | Compare stored episodes against a named baseline with task-specific criteria |
| Should the candidate block a release? | Define an acceptance policy using the platform's evidence and your evaluation process | Run comparison gates and, when validated outcomes exist, outcome-rate gates |
| What does an observed mission success mean? | Check the application's mission or task reporting definition | Check the recorded outcome definition, interval, and coverage limits |
RunFP's regression checks operate on supplied evidence. They do not create new physical trials. Formant's published analytics also distinguishes available data sources; verify that the telemetry you need is accessible in the analysis you plan to run.
Choose a deployment pattern
Keep a fleet platform when your main work is day-to-day robot operations, incident response, and fleet-wide visibility. Use its documented capabilities and confirm the configuration and data retention your team requires.
Add RunFP when you want the documented local workflow for comparing recorded runs, carrying provenance, and applying a repeatable gate. Run provenance and policy-update validation explain the evidence required.
Combine them when field incidents should inform the test suite that approves the next release. Define which system owns incident handling, which owns the reference artifacts, and where the acceptance decision is recorded.
Access to RunFP is arranged through OORB onboarding. Ask each provider about its current commercial terms and deployment options; this page makes no price or infrastructure-cost comparison.
Connect field evidence to a release check
- Identify the task and incident interval in your fleet platform. Preserve the incident or mission identifier and relevant conditions with your experiment.
- Obtain the required recordings through your existing recording or export workflow. Confirm the output format and required signals; an exported CSV or dashboard screenshot is not an MCAP recording. Formant documents export options, but this guide does not assume a native MCAP connector for either platform.
- Obtain an OORB-provided installation, then initialize a workspace and import comparable recordings.
- Choose a healthy reference from the same task and set the named baseline.
runfp baseline set <healthy-episode-id> --name pick-place
runfp check recordings/candidate.mcap --group pick-place --json
- Investigate findings and define valid task outcomes before adopting the result as a release criterion. Keep the fleet incident linked in your team's release record; no automatic Formant or InOrbit synchronization is implied.
Does fleet uptime prove a new robot policy is ready?
Uptime answers an operational question. Policy sign-off also needs candidate identity, task-specific outcomes, comparable conditions, and known coverage gaps. Use the operational metric as part of the evidence and state which release question it supports.
Bring an incident and its run data to an OORB reliability audit.