OhhO Proof
Prove the robot is safe before it ships.
- Category: Trust
- Accent: violet
- Live app: Open validation suite
Run your robot through thousands of simulated scenarios, track regression and coverage, and assemble the evidence into a versioned safety case.
Overview (hero)
Prove the robot is safe before it ships. OhhO Proof runs your robot through thousands of simulated scenarios, tracks regression and coverage, and assembles the evidence into a safety case.
Highlights
- Scenario-based simulation testing
- Regression vs known-good builds
- Coverage map of tested conditions
- Failure / fault injection
- Versioned safety-case report
What you get
- You wouldn't ship software without tests; a robot deserves more. OhhO Proof is the testing and validation product — it puts your robot through scenario-based trials in simulation and on hardware, and tells you, with evidence, whether it's ready.
- Proof runs large batches of randomized scenarios — navigation, manipulation, edge cases and failure injection — across Gazebo and Isaac Sim, scoring each run against safety and task criteria. A coverage map shows what you've actually tested, not what you hope you have.
- Every release is checked for regressions against the last known-good build, and the results roll up into a versioned safety case you can hand to QA, to OhhO Comply for certification, or to a customer who needs assurance.
Features
- Scenario suites — Thousands of randomized navigation and manipulation scenarios across Gazebo and Isaac Sim.
- Pass / fail criteria — Score each run against task success, collisions, safety-zone and timing criteria.
- Regression tracking — Every build is compared to the last known-good; regressions are flagged before release.
- Coverage map — See which conditions — speeds, payloads, lighting, layouts — you've actually exercised.
- Fault injection — Inject sensor dropouts, latency and actuator faults to test failure handling.
- Safety case — Results assemble into a versioned report that feeds OhhO Comply and your QA sign-off.
How it works
- Define scenarios — Pick suites or describe the conditions your robot must handle.
- Run at scale — Proof executes batches across simulators in parallel.
- Read the coverage — See pass rates, regressions and the coverage map.
- Export the evidence — Generate a safety-case report for QA and Comply.
Specs
| Spec | Value |
|---|---|
| Simulators | Gazebo Harmonic + Isaac Sim |
| Scenarios | Navigation, manipulation, edge cases, faults |
| Scoring | Task success, collisions, safety zones, timing |
| Regression | Versus last known-good build |
| Coverage | Condition-space coverage map |
| Output | Versioned safety-case report |
Plans
| Plan | What this product gives you | Included |
|---|---|---|
| Spark | Single-scenario sim tests | ✅ |
| Builder | Scenario suites + regression | ✅ |
| Fleet | Large-batch + coverage + fault injection | ✅ |
| Forge | Custom scenarios + safety-case sign-off | ✅ |
Recommended plan: Fleet. Hobby projects can validate single scenarios on Spark. Teams shipping to real users want Fleet for large-batch testing, coverage and fault injection; safety-critical programs choose Forge for custom scenarios and formal sign-off.
FAQ
Do I need hardware to test? No. Proof runs primarily in simulation, with optional hardware-in-the-loop runs for final validation.
How does it relate to Comply? Proof produces the test evidence; Comply files it as part of the certification record.