Operators were measuring organic wooden housings with digital calipers — 13 dimensions per part, 20 minutes per unit, with measurement variance high enough to undermine SPC. Errors were caught at final assembly, meaning every reject scrapped a unit at its highest point of added value. Previous attempts to automate had stalled. Legacy machine vision required rigid fixturing the line couldn't accommodate. Connecting an AI program to the existing MES and PLCs was a multi-quarter IT effort no one had the bandwidth for. Manufacturing leadership needed a system that worked with the factory as it stood — not one that demanded the factory change first.

Matta deployed in days, not quarters. Standard industrial cameras mounted non-intrusively above the existing workstation, watching parts the way an operator would - no fixturing, no PLC integration, no MES connection. Matta Gauge captured 12 dimensions in 10 seconds against a 3D model reference, with software-defined fixturing replacing rigid jigs entirely. Every measurement flows into Matta as a continuous, part-by-part record. The team gained real-time SPC, predictive alerts when tool wear pushes dimensions toward the tolerance limit, and a queryable history of every part the line has ever produced.

What began as a single pilot station now operates across all production lines. Inspection time fell from 20 minutes to 10 seconds per part at 50μm precision against a 200μm tolerance window — surpassing human gauge R&R standards and eliminating operator subjectivity entirely. The system has run unattended at line speed since go-live. Inspection capacity is no longer a constraint on throughput. The team now uses the continuous measurement record to make upstream tooling and process decisions that were previously made on tribal knowledge.