Skip to content
Planning & productionFor every industry

Analyze downtime causes: why the line stopped

Downtime notes assigned to fixed causes, recurring patterns found, and a clear answer to why OEE dropped.

Today

Downtime gets booked as free text or simply as “other”. The OEE report shows that performance fell, but not why. So every improvement starts with guesswork.

An example

At Talwerk Antriebstechnik, OEE on the geared motor assembly line drops clearly in one week. The agent shows that most of the loss comes from short stops at the screwing station after every type change, on four of five days. It links the alarm logs and two shift log entries as evidence.
Fictional company, typical workflow.

Your data, wherever it lives

Today stops and machine states sit in the MES, alarm logs as exports from the controllers, work orders in the maintenance system, and shift notes as free text in the shift log, on paper or in spreadsheets. demi reads, understands and links them per stop, machine and shift, with your rights. The AI assigns every stop to your cause catalog and shows the biggest losses with evidence in minutes instead of half a day in spreadsheets, and the line lead confirms the classification.

Where the data lives today

  • MES
  • Machines & sensors
  • Tickets
  • 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

    The case needs data from the MES, spreadsheets, shift log and CMMS. demi replaces nothing and ties every stop, with machine states, alarm log, maintenance order and shift note, to machine, shift and order. demi only writes back after approval.

  2. 2 · PermitPermission Layer

    Line and maintenance leads see their own areas, plant management sees the overview. The line lead reviews and confirms the classification, and every change is recorded with a name.

  3. 3 · ActIntelligence Layer

    The agent assigns stops to a fixed reason catalog and traces alarm floods back to the root alarm. It computes OEE from the recorded states, not by estimate, and explains every change through the largest losses and their bookings.

How it runs

  1. 1

    demi reads downtime bookings, alarm logs and shift notes.

  2. 2

    The agent assigns each stop to a reason from your catalog.

  3. 3

    It traces alarm floods back to the first, triggering alarm.

  4. 4

    It finds patterns across shifts, products and machines.

  5. 5

    The line lead reviews the classification and sees the largest losses with reasons.

What changes

  • Real downtime causes in view

  • Targeted improvement instead of guesswork

  • Fewer bookings on “other”

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

+58 hfreed per year

Assumptions and calculation

demi maps downtime notes to fixed causes and finds recurring patterns; you set the actions.

4 h → 2.8 h per case

How does it run today?

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

How we calculate

4 downtime reviews × 4 h × 30% × 1.0 (manual) × 12 months = 58 h

Value per year, before demi costs

Freed per year+58 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

+$3,770

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

Does it need live machine data?

No. demi works with the recorded states and alarm logs that you already have. demi does not touch the machine controls.

Can we use our own reason catalog?

Yes, the agent only assigns within your catalog. If no reason fits, it marks the case for review and can propose a new entry. The line lead decides whether the catalog grows.

How does demi handle alarm floods?

Alarms that arrive in quick succession are grouped into one event, and the agent follows their order back to the triggering alarm. The whole chain stays visible so maintenance can check the result.

How reliable is the classification of free text?

Every classification shows the note it is based on. Uncertain cases are marked for review instead of being filed anyway. Corrections are recorded and serve as examples for later classifications.

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