Skip to content
Planning & productionTextiles

Trace weaving defects back to the loom

demi links every fabric inspection finding to loom, warp beam, yarn lot and shift and shows where a defect is still running.

Today

At fabric inspection every defect is marked and graded, but the link to the loom, the warp beam and the yarn lot gets lost. By the time someone evaluates the inspection sheets, the same loom has woven kilometres more with the same defect. Seconds, rework and complaints cost money, and the cause is argued about in the morning meeting instead of measured.

An example

At Fibrana, start marks pile up in October in fabric GW, 220 g/m² in PA 6.6. The agent maps each inspection finding via its meter position back to loom, warp beam, yarn lot and time: 27 of 34 findings come from loom 14, each right after a stop, and since the control unit was replaced on October 6 the loom's start-up program has been running on factory settings. The other looms on the same yarn lot show no cluster, so the lot is not the cause. The supervisor has the start-up program reset, and the next pieces are clean.
Fictional company, typical workflow.

Your data, wherever it lives

The trail of a weaving defect sits in many sources today: findings on the inspection sheet or in the inspection system, running times and stops in the looms and production data acquisition, warp beams, yarn lots and orders in the ERP, evaluations in spreadsheets. demi reads, understands and links them per piece and meter, with your rights. The AI shows in minutes which loom or lot is causing the defects, and weaving decides what to fix.

Where the data lives today

  • Machines & sensors
  • MES
  • ERP
  • Photos, video & audio
  • Scans & paper
  • Office files
Data LayerOne view of everything, with your rights

How demi solves it

Three layers in this order: connect the data first, then the rights, then the AI.

  1. 1 · ConnectData Layer

    Data sources in this case: looms and production data acquisition, the ERP, the inspection system or inspection sheets, photos and spreadsheets. demi replaces nothing and links every piece with loom, warp beam, yarn lots, article and settings, and maps every meter position back to time, shift and stops via pick density and machine data. Every finding gets its place in the weaving process.

  2. 2 · PermitPermission Layer

    Inspection, weaving and quality see the findings per piece; customer specifications and complaints stay with sales and quality. Looms, warp beams and lots are evaluated, not people. Quality decides on grade, block and customer information, and every decision is recorded.

  3. 3 · ActIntelligence Layer

    The agent assigns each finding to its place in the weaving process with fixed rules, calculates points by your point system and looks for clusters per loom, warp beam, lot, shift and stop. It explains every cluster with the evidence behind it. The language model reads inspectors' notes and voice messages; rules and data do the matching and the points.

How it runs

  1. 1

    Inspection records each defect with type, length and meter position, on screen, by photo of the sheet or by voice.

  2. 2

    demi maps the meter position back to loom, warp beam, yarn lot, time of day and the stops before it.

  3. 3

    The agent grades each piece by your point system and proposes first or second quality.

  4. 4

    It looks for clusters and alerts weaving at once when a defect is still running on a loom.

  5. 5

    The weaving supervisor checks the cause, fixes it and sees in the next pieces whether it worked.

What changes

  • Defects stop at the loom, not at inspection

  • Fewer seconds and fewer complaints

  • Causes based on evidence, not guesses

What does it bring you?

Check it with your own numbers

Every starting value is an assumption you can change. The calculation sits right below, step by step.

Assumption, editable

Your numbers

Result

+216 hfreed per year

Assumptions and calculation

demi traces every weaving defect to loom, warp beam and yarn lot and shows clusters; you fix the cause.

3 h → 1.8 h per case

How does it run today?

What is already digital saves less: the factor lowers the share demi brings.

How we calculate

15 defect analyses × 3 h × 40% × 1.0 (manual) × 12 months = 216 h

Value per year, before demi costs

Freed per year+216 h

Assumption: labor cost of $48.62 per hour in US manufacturing in June 2026 (BLS), plus workplace, IT and overhead.

= Value per year, before demi costs

+$14,040

gained per year

216 h ÷ 1,760 h per role = 0.1 full-time roles

An estimate, not a promise. The starting values are assumptions for an industrial company with 150 to 500 people.

Combine with more cases and get the package

Questions about this use case

How do you get from the meter position to the loom?

Every piece carries loom, warp beam and doffing time. With the pick density and the machine data on running time and stops, demi calculates when each meter was woven. Each finding lands on loom, time, shift and the stops before it.

Does fabric inspection have to work differently?

No. Inspectors mark defects as before. If they write on paper, a photo of the sheet is enough; at the inspection table they can also call out defects, such as a double pick at meter 412. If there is a camera inspection system, demi reads its defect list and images.

Does demi evaluate individual weavers?

No. Looms, warp beams, lots and settings are evaluated, not people. You set how the results are used together with the works council, and demi sticks to those rules.

Which point system does demi grade the fabric by?

The one you use, for example the four-point system, plus the limits a customer sets for an article. Fixed rules calculate the points, each point with its finding. Quality decides on the grade.

See this case with your data

In the demo we show you the workflow on your own systems.

Your configurator

No use cases yet

Your configurator is still empty

Add a use case from any card or case page and see what it brings per year, with the calculation. Free, no sign-up, no commitment: try it out and change any number.

Browse use cases