AI / ML & Computer Vision
CT Garud CT Flow Builder CT OCR / OCV CT Mikha Anomaly Detection BCFHS Vision Measurement Process Check EOL Inspection All vision →
IoT & Smart Factory
Production-Line Tracking EHS / Safety Tracker OEE & Condition Monitoring Digital Dashboards (DMS) NCMS CPAS All IoT →
Safety AI — Enterprise Monitoring Engineering & Machine Building OMR — Exam & Form Scanning Our Work Case Studies About Careers Contact Get Free Assessment →
Book Free Assessment
Vision FMCG AnomalyClassification

Soap Scratch & Dent Inspection

Detects scratches, dents, cracks, and shape deformities on soap bars in real time — classifying defect types so only well-finished bars move forward.

Soap Scratch & Dent Inspection
Real-time
surface grading

/ The Challenge

Surface defects on soap directly hurt shelf appeal, but classification by tired eyes was inconsistent across shifts.

/ What We Deployed

AI surface models detect and classify scratches, dents, cracks, and deformities per bar, rejecting defective bars automatically.

/ The Results

  • Defect types classified, not just flagged
  • Shelf-quality consistency raised
  • Manual inspection effort removed

Built With

Unsupervised AI
Anomaly Detection
Trained on good parts only — flags any surface deviation. No defect labelling. Ideal for castings.
Explore →
Core Vision Engine
CT Garud
AI defect detection & segmentation — 10+ defect types in one frame, <50ms per inference.
Explore →

More in FMCG

Bottle Counting in Crates
Vision FMCG
Bottle Counting in Crates
Counts bottle caps inside crates and boxes automatically, ensuring the correct bottle count before packaging or distribution — short crates rejected instantly.
100% crates verified Case study →
Bundle / Overwrap Count Verification
Vision FMCG
Bundle / Overwrap Count Verification
Counts bottles or products inside each overwrapped bundle — short packs rejected immediately before dispatch.
0 short bundles out Case study →
Extreme Counting — Dense Small Objects
Vision FMCG
Extreme Counting — Dense Small Objects
Counts thousands of small, densely-packed objects in a single image — for packaging and inventory verification where manual counting is impossible.
1000s per image Case study →

/ Get Started

Have a Similar Problem?

Send us your part and inspection challenge. We'll run a free POC and show measured results — no commitment.

Book Free Assessment → See Case Studies