Skip to content
Planning & productionFor every industry

Find the causes of scrap and rework

You see the most frequent causes with evidence, by machine, material, batch and shift.

Today

Scrap and rework are booked with vague reasons and free text. Machine data, batches and suppliers sit in other systems. The real cause stays hidden, and the same defects cost material and time again every week.

An example

At Talwerk Antriebstechnik, scrap on ZR-3140 gears on line 2 doubled in March. The agent shows that 31 of 38 cases share one steel batch from a new supplier and only started after a change to the hardening parameters. The quality lead puts the remaining stock on hold and has the batch tested.
Fictional company, typical workflow.

Your data, wherever it lives

Scrap is booked with free text in the MES and ERP today, process parameters live in the machine controls, batches and suppliers in the ERP, inspection results in the QM system, rework in spreadsheets and hints in the handwritten shift log. demi reads, understands and links them per part, machine, batch and shift, with your rights. The AI shows patterns in minutes that take hours by hand, and the quality lead decides on the action.

Where the data lives today

  • MES
  • ERP
  • Machines & sensors
  • QM system
  • Office files
  • Scans & paper
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

    The case needs data from the MES, ERP, QMS and spreadsheets. demi replaces nothing and ties every scrap and rework report to part, order, machine, material batch, supplier and shift. Agreed actions go back only after your approval.

  2. 2 · PermitPermission Layer

    Quality and production leads see the analysis, which runs by line, machine and shift, never by name. The quality lead approves actions, and this is recorded.

  3. 3 · ActIntelligence Layer

    The agent maps free-text reasons to fixed defect types and looks for links with batches, machines and shifts. Quantities and costs come from the bookings, not from estimates, and every link shows the records behind it.

How it runs

  1. 1

    Read in scrap bookings and free-text reasons

  2. 2

    Map free text to fixed defect types

  3. 3

    Link with machine, batch, supplier and shift

  4. 4

    Show the most frequent causes with evidence

  5. 5

    Set up an action and track its effect in the following weeks

What changes

  • Causes found faster

  • Lower cost of scrap and rework

  • Improvement projects based on data

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

+72 hfreed per year

Assumptions and calculation

demi links scrap to machine, material, batch and shift and shows the causes; you fix them.

5 h → 3.5 h per case

How does it run today?

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

How we calculate

4 root-cause reviews × 5 h × 30% × 1.0 (manual) × 12 months = 72 h

Value per year, before demi costs

Freed per year+72 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

+$4,680

gained per year

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

Can demi read free-text scrap reasons?

Yes. The agent maps free text such as “oversize, re-measured” to the defect types you define. Uncertain matches are flagged for review. The original text stays visible on every report.

How does it link scrap to material batches?

Through the production order: the ERP or MES holds which batch it consumed. demi joins that information with goods receipt, supplier and certificate. You see whether defects cluster around one batch or one supplier.

Are individual operators identified?

No. The analysis runs by line, machine, shift and batch. Names are not shown, even when they appear in the bookings. This is set as a fixed rule in the Permission Layer.

Is a link the agent finds already the cause?

Not always. The agent shows clusters and the records behind them; your team makes the call. Often it is the starting point for a targeted test or an 8D report.

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