Back to homeDeutsch

by

Agents that work alongside
you in your factory.

AI agents read purchase orders, check the details and show where your orders stand. Your people make the decisions.

the path of an order
The path of an order

Schematic view

Purchase order comes in
Order
from the purchase order
reads purchase orders
Planning
Material, due dates
calculates suggestionsin development
Production
Instructions at the workstation
reads work instructions
Shipping
Goods go out
Delivery goes out
ERP and MES stay as they are

The path of an order and the points where an agent helps.

20 hours → 2

At a CNC contract manufacturer, the weekly effort fell from 20 to 2 hours. Today, people check the exceptions.

90 percent automated in the published CNC case · published case study

what changes

The same people, different work.

Your people check the results. The application takes over repetitive entry steps.

Typing purchase orders line by line
→
Approving deviations
Copying work instructions from PDF
→
Checking flagged values
Asking around the shop where an order is
→
Looking at the screen
running today

See orders where the work happens.

Ten screens show orders, due dates and deviations, one per department.

Department screen in the DUAP turning shop showing current orders, due dates and deviations.
DUAP · Ten screens in around 30 hours of engineering effortProduction order numbers, parts and KPIs blurred
running today

Check work instructions instead of retyping them.

The review screen shows the PDF and the extracted fields side by side. Uncertain values are flagged.

Review screen: the work instruction PDF on the left, the extracted fields with a confidence rating on the right.
GEWO · In the example shown: 32 of 36 fields with high confidence.Sample document, no customer data
running today

A person chooses what the model learns from.

Each line is one work cycle of a motor. A person excludes outliers before the model trains.

Current curves of many motor cycles overlaid, below them a switch for each cycle, some of them excluded.
maxon · running in production in the maxon AzureNames and numbers blurred
in development

Review suggestions, then confirm.

Stock and capacity appear as a six-month preview. The planner confirms the change.

Production planning with parts over six months, bottlenecks highlighted in colour.
In development, not yet in use at customers · invented sample data
and my data?

You decide where it runs.

The same factory three times. The only thing that moves is the green dot.

DATA STAYS ON SITE

On your premises

Data processing can run in your plant. At DUAP, the data stays on site.

IOSYS · CH/EU

Operated by us

iosys handles operation and updates on servers in Switzerland or the EU.

YOUR MICROSOFT CLOUD

In your Azure

The application runs in your Microsoft environment. This is how maxon runs MIND today.

The green dot shows the intended processing location.

references from iosys projects

From iosys projects.

maxon · DUAP · GEWO · UNIFIL · Eglin · SMPtec

These companies stand for earlier iosys projects, not for use of factorIQ.

get started

Where does your factory start?

Usually with the module that costs the most time today. Tell us which repetitive work that is for you. We will tell you whether it is worth it.

Arrange a call →

factorIQ by iosys · iosys.swiss