⚡ AI Surface Defect Detection
Live detection view — scratches, dents, scale and pinholes are located and labeled on the strip surface in real time.
AI‑driven machine vision continuously scans strip surfaces, detecting and classifying defects in real time — turning surface quality from sampling checks into 100% inline coverage.
Deep‑learning models trained on real mill data recognize and localize the most common strip surface defects with confidence scores.
The system detects, localizes and grades defects across the full strip width and length, building a complete defect map for every coil.
Steel surface defect detection has relied on human visual inspectors for over a century — and the results have been consistently inadequate. A trained inspector examining hot‑rolled coils on a finishing line at production speed sees roughly 60–70% of surface defects present on the strip. The remaining 30–40% pass through undetected — scratches masked by scale, edge cracks hidden by strip curvature, inclusions too subtle for the human eye at line speed, and lamination defects invisible on the surface until downstream processing reveals them.
A flat‑rolled steel producer shipped 14,200 tonnes of coil to an automotive stamping customer over six months before a pattern of press‑shop cracking revealed a systematic surface inclusion defect that had been present since the caster. The root cause was a tundish nozzle erosion issue that created alumina streaks on the slab surface — defects that were technically visible on the hot strip mill exit but occurred at a frequency and contrast level that human inspectors could not reliably detect at 900 metres per minute. The total cost of the quality escape exceeded $3.8 million in customer claims, returned material, re‑inspection labour, and lost future orders from a tier‑one automotive account that took eighteen months to recover.
AI‑powered vision inspection systems have fundamentally changed what is possible in steel surface defect detection. Deep learning models trained on millions of labelled defect images can now detect, classify, and grade surface defects at production speed with accuracy levels exceeding 95% — operating 24 hours a day without fatigue, inconsistency, or the subjectivity that makes human inspection unreliable. These systems identify defect types that human inspectors cannot see at line speed, correlate defect patterns back to upstream process conditions, and feed real‑time quality data into CMMS and production systems to trigger immediate corrective actions.
SWOTY integrates AI vision inspection data directly into maintenance and quality workflows — connecting surface defect detection to root cause analysis, equipment condition tracking, and corrective action management across the entire steelmaking process chain. Start your free trial to see how AI vision transforms steel quality management from reactive customer complaints to proactive defect elimination at the source.