Facade errors are cheap in the model and ruinous on site. The gap between the two is an audit nobody scheduled.

What the audit catches

Before anything is cut, we check three things the model tends to hide: real-world tolerance stack-up across panel joints, the actual unique-panel count versus what the budget assumed, and whether every panel still hosts cleanly when the geometry shifts.

None of these show up as a red flag in Rhino. They show up as a delivery that does not fit.

A facade that closes in the model can still miss by a panel on site. The audit is where you find out which one.

The output is a short report the project team can act on while changes are still a slider, not a variation.