Optical components that survive automotive validation.

Birefringence under thermal load shows up at the contrast test, not the caliper.

What we do about it: model retardation across the operating range during prescription review, then verify on coupons from the actual press and resin lot before tool kickoff. PPAP-dimensional parts that lose contrast at 70 °C are not parts that ship.

Alignment drift across thermal cycle invalidates the calibration the system was qualified at.

What we do about it: lock datums to the optical axis at design freeze, gauge that the molded part holds those datums across the spec'd thermal envelope, and write the gauge R&R that captures it so the supplier handoff is unambiguous.

Residual stress at the gate creates a clear-aperture hot spot no first-article will catch.

What we do about it: route gates with mold-flow that respects the clear aperture, then sample under crossed polarizers on the production cavity, not a prototype cavity. Stress witnesses are part of the lot release, not an audit finding.

Environmental exposure — humidity, UV, salt — drives haze and coating shift that pass at week one and fail at month six.

What we do about it: bake AEC-Q104-style environmental coupons into the qualification plan, with adhesion and transmission deltas captured against the as-deposited stack so the lifetime claim has data behind it.

PPAP confirms dimensions and packaging — not optical performance.

What we do about it: add wavefront, transmitted MTF, scatter, and thermal-cycle results to the release record alongside CMM. The part is qualified for what the system actually does, not just for what the print calls out.