Textiles · TX-05 · AI fabric inspection

Defects must be found before cutting.

Garment manufacturers worldwide are replacing manual fabric inspection with line-scan camera and AI systems. The defect map is passed through to the cutting stage. Fabric production in Uzbekistan is growing rapidly.

Platform: Plant Pulse

94.2 mn m²

Cotton fabric output, January–September 2026 (2025: 72.5 mn m²).1

20+

Defect categories detected by EagleAi, according to OSHIMA.2

90%+

AI detection rate according to the vendor (manual: 60–70%).2

A roll defect that reaches a cut part is costly.

Who it is for

Head of Quality
Defect statistics.
Cutting room
Fabric utilisation.
Procurement
Supplier claims.
CEO
Returns.

Where the problem lies

  • Sample inspection

    Only part of each roll is inspected.

  • No defect map

    Defect locations do not reach cutting.

  • Supplier

    Defective roll statistics are not collected for supplier claims.

  • Shade variation

    Shade variation between lots is detected late.

How the loss is calculated

Cut partsdefective cut parts per month × cost per part
Returnsreturned orders per month × order value
Claimsvalue of defective rolls not returned to the supplier per month
Inspectioninspection hours per month × number of inspectors × hourly rate

Calculate with your data

Cost of Cut Pieces Scrapped Due to Fabric Defects

This calculator estimates one type of loss per month.

Use spaces for thousands and a dot or comma for decimals. Use the same currency throughout.

Values stay in this page and are not sent to the server.

Scenario estimate

0 / 4 values entered

Enter your operating figures to see the estimate.

The results are estimates, not promises, and are only as reliable as the values you enter. Enter the cost basis per piece, such as fabric, cutting and other costs incurred, not the selling price. The results do not predict the detection rate of any AI equipment. They do not guarantee savings, compliance with buyer requirements or certification. A reliable assessment requires your own measured data and an audit.

Formulas and assumptions

This calculator estimates one type of loss per month. It covers the cost basis of cut pieces scrapped because a fabric defect was not detected before cutting, minus any value recovered through rework or salvage. Enter the current monthly piece count from your quality records. The target count is also your own figure, and the calculator does not claim the target is achievable. Customer returns, supplier claims, inspection labour, investment and payback are excluded.

Current monthly loss
Current scrapped pieces (pieces/month) × (cost basis per piece − value recovered per piece, currency/piece). A negative difference is treated as zero.
Monthly loss in target scenario
Your target scrapped pieces (pieces/month) × (cost basis per piece − value recovered per piece, currency/piece). This is not a guaranteed outcome.
Monthly difference between current and target
(Current − target scrapped pieces, pieces/month) × (cost basis per piece − value recovered per piece, currency/piece). Negative values are set to zero. This shows estimated loss exposure, not confirmed savings.

Facts and global practice

  • National Statistics Committee

    Cotton fabric output in January–September 2026 was 94.2 mn m², compared with 72.5 mn m² last year.1

  • OSHIMA press release

    OSHIMA EagleAi detects, maps and classifies more than 20 defect categories in woven, knitted and elastic fabrics.2

  • OSHIMA press release

    Defect data is sent to a cloud platform and projected onto the fabric before cutting; the system integrates with MES and ERP.2

  • Briefglance (on OSHIMA)

    Data cited by Briefglance: AI detection above 90% versus 60–70% for manual inspection; inspection 6 times faster.2

The detection rate is a vendor claim. Model testing is required for each fabric type.

Defect data travels with the roll

AI inspection equipment is supplied by a partner. MEENTAKA links defect data to the roll, warehouse and cutting.

  1. 01Inspection
    • Camera
    • AI
    • Defect category
  2. 02Roll
    • Lot
    • Storage location
    • Grade
  3. 03Cutting
    • Map
    • Fabric-saving nesting
    • Cut parts
  4. 04Report
    • Supplier
    • Claim
    • Statistics

Modules

The current maturity level is shown next to each module.

Roll warehouse (WMS)Lot and locationCommercially available
Linking defect dataFrom partner equipmentIn development
Supplier ratingFor claimsEngineering capability
AI inspection equipmentPartnerIn development

How we start

  1. 01

    Quality audit

    Inspection process.

    Outcome: defect statistics.

  2. 02

    Pilot with partner

    One fabric type.

    Outcome: detection rate.

  3. 03

    Integration

    Warehouse and cutting.

    Outcome: full chain.

  4. 04

    Subscription

    Monthly report.

    Outcome: trend.

What is measured

Defective rollsShareSystem
Defects at cuttingCut partsControl
ClaimsAmountProcurement
Inspection timePer rollSystem

What is needed to start

  • Head of Quality
  • List of fabric types
  • List of suppliers
  • Warehouse layout

What we do not promise

  • Exact timeline and price without an audit
  • Guarantee of meeting buyer requirements
  • Certification
  • Unmeasured results or savings
  • Detection rate of the AI equipment

Self-check

  1. Is every roll fully inspected?
  2. Are defect locations recorded?
  3. Does defect data reach cutting?
  4. Are suppliers rated on defects?
  5. Is the stage at which shade variation is detected known?
  6. Is roll grade recorded in the system?
  7. Are claim statistics kept?
  8. Is the number of inspectors sufficient?
  9. Are the causes of returned orders analysed?
  10. Is there a designated responsible person?

Quality audit

We review fabric inspection and roll records and show where defects are slipping through.

  • Defect statistics
  • Partner pilot plan
  • Implementation fee and monthly subscription

Sources

  1. National Statistics Committee: industrial production press release, January–September 2026. stat.uz, 30.10.2026.
  2. OSHIMA press release: EagleAi fabric inspection (Texprocess 2026). taiwannews.com.tw, 30.04.2026; briefglance.com.