{"stable_id":"8ed0ede9e7831b2f","slug":"audi-weld-quality-ai","company":"Audi Neckarsulm","workflow_name":"WPS Analytics full-population inspection","function_code":"engineering","pattern_codes":["threshold_as_control","exception_based_operations"],"changed_assumption":"Quality can move from ultrasound sampling to AI screening of every weld.","evidence_strength":"reported","publication_tier":"showcase","freshness":"current","reviewed_at":"2026-08-23","updated_at":"2026-08-22","source_quality_summary":"1 independent, 1 primary; publication outcomes are reported.","caveat_summary":"Operator coverage, not defect escape-rate evidence.","freshness_caveat":null,"workflow_summary":{"before":[{"actor":"Audi quality technician","action":"Uses ultrasound to inspect a random sample of resistance spot welds.","actor_type":"control"},{"actor":"Quality engineer","action":"Reviews sampled findings and directs investigation or rework.","actor_type":"human"}],"hinge":"Quality can move from ultrasound sampling to AI screening of every weld.","after":[{"actor":"WPS Analytics","action":"Analyzes process data for approximately 1.5 million welds across 300 vehicles each shift.","actor_type":"system"},{"actor":"Quality technician and engineer","action":"Focus on flagged anomalies and continue random ultrasound validation.","actor_type":"human"}],"decision_mode":"shared","decision_marker":"Human sets the threshold; AI decides each instance"},"before":[{"order":1,"actor":"Audi quality technician","action":"Uses ultrasound to inspect a random sample of resistance spot welds.","handoff_to":"Quality engineer","control":"About 5,000 weld checks per vehicle reported; sample-based control remains in parallel."},{"order":2,"actor":"Quality engineer","action":"Reviews sampled findings and directs investigation or rework.","handoff_to":"Body-shop production","control":"Audi quality procedures and human disposition."}],"after":[{"order":1,"actor":"WPS Analytics","action":"Analyzes process data for approximately 1.5 million welds across 300 vehicles each shift.","handoff_to":"Quality dashboard anomaly queue","control":"Machine-learning screening covers the population; traceability designed with quality/Fraunhofer bodies."},{"order":2,"actor":"Quality technician and engineer","action":"Focus on flagged anomalies and continue random ultrasound validation.","handoff_to":"Rework or release decision","control":"Humans own disposition; full coverage is not proof of zero defects."}],"decision_rights":"Audi sets thresholds; humans disposition anomalies.","exception_path":"Flagged or uncertain welds receive manual ultrasound review.","removed_work":["Routine review of non-anomalous welds"],"outcomes":[{"metric":"Inspection coverage","baseline":"Random checks around 5,000 welds per vehicle","result":"AI analyzes about 1.5 million welds on 300 vehicles each shift","period":"2023 rollout","scale":"300 vehicles per shift","attribution_caveat":"Operator coverage, not defect escape-rate evidence.","evidence_label":"reported"}],"executive_lesson":"Skilled inspection shifts from broad sampling to anomaly disposition.","anti_pattern":"Do not infer zero defects from 100% coverage.","questions_for_leaders":["Where is the operating threshold set and who can override it?","What measured result would trigger rollback or retraining?","Which residual decisions must remain human-owned?"],"collections":["human-still-decides","embodied-work"],"bundle_version":"1.0.0","bundle_fingerprint":"sha256:c23c6cc2b88153d008ea8fda928f632ce0011fc2d4c5036672a16e5d895bab93","canonical_url":"https://brianletort.ai/transformations/audi-weld-quality-ai"}