Skip to content
Quality, standards & complianceElectronics & SensorsMetal & MachiningPlastics & PackagingAutomotive Suppliers+3

Limit samples always at hand

The inspector photographs the defect, and demi shows the matching limit samples, customer agreements and earlier decisions.

Today

Whether a black speck, a flow line or a burr is still acceptable is judged differently by every shift. The limit samples sit in a cabinet in the inspection lab, agreements with the customer in old e-mails and visit reports, earlier decisions in the heads of experienced inspectors. At night, doubtful parts are scrapped or waved through, and both cost money: good parts in the scrap bin or a complaint from the customer.

An example

At Polivo Kunststoffverpackungen in Weißwasser, the night-shift inspector finds a dark speck in the base of a 500 ml BH container made of rPET. She photographs it at the inspection station; the agent measures 0.6 mm and shows the two matching limit samples: for the dairy customer, specks up to 0.8 mm outside the neck area are acceptable, as recorded in the minutes of a customer visit in March. It also shows a decision from three weeks ago, when 14 cases with similar specks were released. The inspector releases the batch instead of blocking 9,600 containers, and the agent files the photo as a new borderline case for the quality manager.
Fictional company, typical workflow.

Your data, wherever it lives

Limit samples sit today as parts in a cabinet in the inspection lab, photos of them scattered across the network drive, customer agreements in e-mails and visit reports, deviation releases and complaints in the QM system and the DMS. demi reads, understands and links them, with your rights. The AI shows the matching limit sample with its agreement at the inspection station in seconds, and your inspectors decide.

Where the data lives today

  • Photos, video & audio
  • QM system
  • E-mails
  • DMS & archive
  • PDFs
  • Folders & drives
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 QM system, the DMS, e-mail, inspection records and photos. demi replaces nothing and links every limit sample with photos under defined lighting, defect type, inspection zone, article, customer and agreement, plus earlier blocks, deviation releases and complaints. demi writes new decisions into the catalogue only after approval.

  2. 2 · PermitPermission Layer

    Inspectors and shift leaders see the limit samples of their articles with the released agreement behind them; full visit reports, e-mails and complaint details are visible only to quality assurance. A new or changed limit sample applies only once quality management releases it, and where the customer must confirm it, only after the customer does. Every decision is recorded with photo, person and time.

  3. 3 · ActIntelligence Layer

    The agent recognises defect type, position and size on the photo, compares them with the limit sample photos of the same article and shows the closest ones, with the agreement and the decisions behind them. What can be measured, such as the diameter of a speck or the length of a scratch, it measures on the image with a scale; what is acceptable it takes from the released catalogue. It proposes a classification; a person decides whether the part is OK.

How it runs

  1. 1

    The inspector photographs the suspect part at the inspection station, under the set lighting and with a scale.

  2. 2

    The agent recognises defect type, position and size and finds the matching limit samples for this article.

  3. 3

    It shows the limit sample photos, the customer agreement and earlier decisions on similar cases.

  4. 4

    The inspector decides; in case of doubt, one click brings in quality assurance, who see the same image.

  5. 5

    Every decision goes into the catalogue with its photo, and the agent proposes new borderline cases as limit samples.

What changes

  • Every shift judges cosmetic defects the same way

  • Fewer good parts scrapped and fewer complaints

  • Experienced inspectors' knowledge stays in the catalogue

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

+108 hfreed per year

Assumptions and calculation

demi shows matching limit samples and past decisions for every photo; the inspector decides.

15 min → 10.5 min per case

How does it run today?

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

How we calculate

120 borderline cases × 15 min × 30% × 1.0 (manual) × 12 months = 6,480 min = 108 h

Value per year, before demi costs

Freed per year+108 h

Assumption: labour cost of €49.50 per hour in German manufacturing in 2025 (Destatis), plus workplace, IT and overheads.

= Value per year, before demi costs

+€7,020

gained per year

108 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 our physical limit samples get into demi?

Each limit sample is photographed once, under set lighting, with a scale and from the agreed angles. The agent assigns the photos to article, defect type and inspection zone and links them with the agreement behind them. The physical sample remains the reference; the photo makes it available at every inspection station.

Does demi decide whether a part is OK?

No. The agent proposes which limit sample a defect comes closest to and shows the evidence. Inspectors and quality assurance decide, and for a new borderline case quality management decides whether it becomes a limit sample.

What about agreements that exist only in e-mails?

That is exactly what demi finds. The agent reads e-mails, visit reports and deviation releases, recognises agreements on defect types and zones and puts them next to the matching limit sample. Quality assurance confirms that an agreement still applies before it goes into the catalogue.

Does this also work for paint, wood or textiles?

Yes. Whether it is an inclusion in paint, a knot in veneer, a weaving fault or a scratch on anodised aluminium, limit samples exist wherever appearance matters. What counts is that photos are always taken under the lighting that the visual inspection prescribes.

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