Today
Spare-part stores hold thousands of parts for old machines and tools, many for years, booked as a reserve in case a customer still orders. Other parts cannot be bought at all any more, and remaking them takes weeks. Metal and polymer printing could supply many of them on demand, but nobody checks part by part which ones are suitable: material, size, geometry, tolerances, demand and storage cost are spread across CAD, PLM, ERP and stock lists.
An example
Norvelta holds about 900 spare inserts, cores and slides for 240 customer moulds, some of them for more than ten years. The agent checks every part against model, drawing and withdrawal history and finds 86 candidates: inserts in 1.2343 that fit the build volume of a metal printer, whose fitting faces can be ground or spark-eroded after printing and that were called off at most twice in ten years. 23 of them have cooling bores that could even become conformal cooling channels once printed. For each part it shows storage cost, past manufacturing time and the finishing after printing. The head of tool service releases the first 30 parts and agrees the switch to a printable tool steel with each customer.
Your data, wherever it lives
What sits in your spare-part store is spread across systems today: 3D models in the CAD system, drawings with tolerances in the PLM or as PDFs, stock, bin locations and withdrawals in the ERP, and spare-part agreements with customers in spreadsheets. demi reads, understands and links them per part, with your rights. The AI checks thousands of parts for printability in days, work that would otherwise take months of opening models, and your service management decides what gets printed in future.
Where the data lives today
- CAD & drawings
- PLM
- PDFs
- ERP
- Office files
How demi solves it
Three layers in this order: connect the data first, then the rights, then the AI.
- 1 · ConnectData Layer
Data sources in this case: 3D models in the CAD system, drawings with tolerances and surface finishes in the PLM or as PDFs, material master, stock, bin locations and withdrawal history in the ERP, and spreadsheets with customers' spare-part agreements. demi replaces nothing and links every spare part to its model, drawing, material, withdrawals and storage cost. Each part gets a record from which its printability can be read.
- 2 · PermitPermission Layer
Design and service see models, drawings and stock; costs and customer agreements are visible to service and plant management. Data on a customer's tool is visible only to the team serving that customer. An authorised person releases which parts will be printed in future, and that is recorded.
- 3 · ActIntelligence Layer
demi computes dimensions and volume exactly from the 3D model, along with wall thicknesses and internal channels, and matches them against the build volume and materials of your printers or print service. The agent reads tolerances and surface finishes from the drawings and marks which faces must be machined after printing. It ranks the candidates by criteria you set: storage cost, demand frequency, lead time of a remake.
How it runs
- 1
demi reads the spare parts in stock with model, drawing, material, stock level and withdrawals.
- 2
Every part is checked against build volume, printable materials, wall thicknesses and tolerances.
- 3
The agent ranks the suitable parts by storage cost, demand frequency and the lead time of a remake.
- 4
For each candidate it shows what must be machined after printing and whether a material change needs approval.
- 5
Service management selects the parts to be printed on demand and releases stock reduction and print data.
What changes
Less capital and floor space tied up in spare parts
Even parts you can no longer buy are available within days
A checked list of printable parts instead of one-off decisions in an emergency
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
+96 hfreed per year
Assumptions and calculation
demi checks material, geometry, demand and storage cost per spare part; you decide what gets printed.
30 min → 18 min per case
How we calculate
40 spare parts × 30 min × 40% × 1.0 (manual) × 12 months = 5,760 min = 96 h
Value per year, before demi costs
Freed per year+96 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
+€6,240
gained per year
96 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 packageQuestions about this use case
Which parts are suitable for printing at all?
That depends on material, size, geometry, tolerances and quantity. Parts with complex shapes, small quantities and long lead times are often well suited; large, simple turned parts or materials that cannot be printed are not. You set the rules with your production or your print service, and demi applies them to every part.
What about tight tolerances and fitting faces?
The agent reads them from the drawing and marks every face that must be ground, milled or spark-eroded after printing. A part counts as a candidate only if that finishing is feasible. The rating shows its reasons so your production can check it.
May a printed part replace the original material?
You decide, often together with your customer. The agent shows where the printable material differs from the original, for example in hardness or thermal conductivity, and marks every change that needs approval. Without approval the part stays as it is.
What happens to the parts in stock?
Nothing without your decision. demi proposes which stock can be reduced and which parts will in future be kept only as checked print data. Scrapping, rebooking and agreement with the customer are released by an authorised person.
