Today
Capacity planning counts machines, but not who is actually present and qualified. Bottlenecks only show up on the day. Overtime and moving people around are then decided on the spot.
An example
At Talwerk Antriebstechnik, the welding cell is well over capacity in week 14 because two certified welders are on leave. The agent proposes moving one qualified colleague from the night shift to the early shift for three days and shows the effect on the night shift at the same time. The shift lead confirms the proposal.
Your data, wherever it lives
Orders and routings sit today in the ERP, machine times and confirmations in the MES, shift plans and the qualification matrix in spreadsheets, absences in the HR system and leave calendars. demi reads, understands and links them per work center, week and person, with your rights: absences count only as availability, never with a reason. The AI sees bottlenecks weeks ahead instead of on the day and runs scenarios in minutes, and the shift lead decides who works where.
Where the data lives today
- ERP
- MES
- HR system
- Office files
- Calendars
How demi solves it
Three layers in this order: connect the data first, then the rights, then the AI.
- 1 · ConnectData Layer
The case needs data from the ERP, MES and spreadsheets. demi replaces nothing and links production orders and routings with machine calendars, shift plans, absences and the qualification matrix. demi writes the shift lead's decision back only after approval.
- 2 · PermitPermission Layer
Absences appear only as availability, never with a reason. The shift lead decides on staffing, and every change to the plan is recorded.
- 3 · ActIntelligence Layer
The agent computes load against capacity per work center and week exactly from routings and calendars and shows the effect of absences and overtime. It proposes staff allocation by qualification, each bottleneck with the orders behind it.
How it runs
- 1
Read in orders, routings and calendars for the coming weeks
- 2
Add the shift plan, absences and qualifications
- 3
Compute load against capacity per work center
- 4
Show bottlenecks and run what-if scenarios
- 5
Propose to the shift lead; the decision goes back into the shift plan
What changes
Bottlenecks visible in advance
Qualified staff used better
Planned rather than ad-hoc overtime
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.
Your numbers
Result
+58 hfreed per year
Assumptions and calculation
demi compares the weekly load with machines and staff, absences included; you adjust.
4 h → 2.8 h per case
How we calculate
4 capacity 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 packageQuestions about this use case
Does demi use our shift and absence data?
Yes, from time recording, the HR system or shift plans in spreadsheets. Planning only needs availability: present or not. The agent never sees the reasons for an absence.
How are qualifications handled?
Through the qualification matrix: who may run which machine or process, including certificate validity. A proposal only uses people with a valid qualification. If one is about to expire, you see it in the plan.
Can I try out overtime or an extra shift?
Yes. You create a scenario, such as a Saturday shift or a move between teams, and see the effect on every work center. The current plan stays unchanged until the shift lead decides.
Does demi also plan the sequence at each machine?
This use case covers load and capacity per week and work center. The sequence at a single machine is a separate use case that builds on the same data.
