Skip to content
onagero

AI in the Foundry: From the Melt Log to the Cause of Scrap

In a foundry, a defect often shows only days after casting, during fettling, inspection or machining. The link back to the melt and the mold is then laborious. Onagero establishes it from your data.

  • Runs on your server
  • Reads only, never writes
  • GDPR ready
  • First result in weeks

In short

With Onagero, foundries build their own tools that connect scrap and complaints with melt, batch, pattern and shift, show melting and material costs per casting, and assemble times across molding, casting, fettling and machining. Read-only, on your server.

What is different in Foundry

The defect shows up late

Days and several stations lie between casting and finding. The tool traces a casting back across these stations, as far as the data allows.

Returned material

Risers, gates and scrap go back into the melt. Where input quantities are booked, the yield per part becomes visible.

Energy is a cost block

Where melting energy is recorded per batch, it can be included in the cost per casting.

Your systems

Foundries often have an ERP, melt logs in a separate application or in Excel, and inspection data in the CAQ. Onagero connects them via batch, order and time.

How the connection works

The data path at Onagero Your systems feed, over a read only connection, into a masked analysis copy on your server. Your people build internal tools from it with Onagero. Only the released, masked parts of a request go to the model provider. Your server, your network ERP BDE and MES Excel, files Analysis copyNames masked OnageroYour people build Internal toolwith roles per person Live view, only what thetool may read Model provider, external What goes out is only what a request needs: released and masked. The databasestays in your building. The data path at Onagero, top to bottom Your server, your network ERP BDE, MES Excel Analysis copyNames masked OnageroYour people build Live view, released Internal toolwith roles per person Model provider, external Only what a request needs, released and masked. The database never leaves your building.
The path of your data: platform, analysis copy and tools on your server; only masked parts of single requests go to the language model. Data security in detail
Two colleagues walk through a process on paper on the shop floor.

Start with one problem.
Keep the capability.

No AI transformation. One real process, solved on your data in a few weeks. Your team solves the next one itself.

Questions from Foundry

Can Onagero control the melt?

No. Onagero reads logs and bookings. It does not control furnaces and does not suggest alloy additions.

What if castings cannot be traced individually?

Then the tool works at batch or day level and says so. Finer traceability is a question of marking, not of software.

Are X-ray or inspection images included?

Not in the standard scope. Onagero works with the findings that are booked.

How do we begin?

With the process that ties up the most time, often the scrap evaluation. The fit check helps with the selection.

Which problem would
your team solve first?

Show us a real process. We will show you how it could look with Onagero.

Book a demo →

Check first whether a pilot fits: fit check in 2 minutes →