The same repo fixed the opposite error on the same scan, a false Approximation warning from a fit that chased scan noise. One scan produced both a missed warning and a false one.
Recurring agent failure issue in AeroCAD (FlappedEar)
What engineers report
A recurring agent failure problem in AeroCAD (FlappedEar) that mostly hits the hardware engineer.
Failure pattern: validation / failure
Analyst note
From the Grok pain-radar digests, Nov 8-9 2026 (Nov 8 8:41 AM ET through Nov 9 6:41 AM ET digests plus the Haitch editions). Verified live 2026-11-09: issue/PR read; title and state confirmed.
What it costs a team
Modeled annual cost
$30,000
For a team of 5
Prototype delay
3 days
Modeled schedule slip per occurrence
Time lost
240 min
Per engineer per month, about 1 incidents
How the cost is modeled
240 min/month × 5 engineers × 12 months = 240 hours × $125/hour = $30,000
A modeled estimate with medium confidence. Swap in your own team size and loaded rate to size it for your team.
Engineering tools that address it
No engineering tool profile lists this error yet. Browse tools that connect to AeroCAD (FlappedEar).
Evidence
More recurring errors in AeroCAD (FlappedEar)
- A remembered cm/in/m import setting was applied again after the STEP, IGES and 3MF readers had already converted to millimetres, so parts came in at the wrong size. The fix applies embedded units once and shows unit provenance in the import status.MCAD · impact 74
- A construction plane was meshed as a solid wall in the CFD run and skewed lift and side force: with Other visible parts in the flow on, a reference plane went into context.stl and snappyHexMesh meshed it as a thin wall upstream of the part. Lift read -0.261 N with the plane and -0.155 N without.Verification · impact 66
- A fillet on a fin edge down to a scan root reported OK, but the export check then refused print export with 3 to 78 open edges, and some edges failed the B-rep check. Meshing the live shape left mismatched triangulation and changed tolerances that undo still held. The fix makes a fillet count as failed unless the export-resolution mesh is closed.MCAD · impact 66
- An open scan shell with 329 open edges went into a CFD run with no warning, and snappyHexMesh cannot tell inside from outside, so the forces depended on the mesh rather than the part. Measured on OpenFOAM, open shell drag was 1.83 N against 0.82 N after repair, and the fix is a warning rather than a refusal.Verification · impact 66
- In AeroCAD, an STL in mm imported with unitless files set to m came in 234 m long. The size check named the wrong unit even though the factor was right. The import unit was not stored on the part, so the check guessed from the numbers.MCAD · impact 66
- STEP export dropped body names and colours, and import merged bodies. Exported bodies came out with OCCT placeholder names and no colours, and a two-body STEP imported as one part named after the file. Fourth STEP handoff bug in today's sweeps.MCAD · impact 66
Sourced from public community reports. Costs and delays are modeled estimates, not measurements. Know a better source or a fix? Tell us at hello@productflo.io.