Case - Auto parts | Packing miss-build detection

Customer background

A brake-system plant supplying global OEMs. Lines assemble calipers, pads and kits, then verify missing parts, layer count and pieces per layer. IATF demands traceability and packing compliance.

Missing parts used to show up downstream or at the customer. Packing progress lived on paper with no process proof. A miss-build can fail braking, trigger recall, and paper checks cannot keep a piece-level record. The plant needed AI on assembly and packing.

Pain points

Missing assembly is found too late

Many parts and steps. Human self-check fatigues. Missing bolts are found after they leave, and rework is costly.

Packing progress is logged by hand

Counts live on paper. Disputes lack inspectable screenshots and records.

IATF traceability is strict

Packing compliance and piece-level proof are hard to meet with error-prone manual checks.

Solution

AI poka-yoke for assembly and packing: missing-part detection, part type, layer count, pieces per layer and case ID bind. Live miss-build catch and packing checks, 7×24, across mixed brake SKUs.

Helmet / glove

Zone intrusion alert

People-vehicle distance

Work-at-height PPE

Climbing hazard

People-equipment distance

Miss-builds are stopped at the station. Cases, layers and counts are logged with screenshots. Dashboards show pass rate, hot stations and operators, and IATF-ready reports. The same method transfers to energy storage and other high-consistency industries.