// approach

A path that de-risks the decision

You shouldn’t have to bet on a system that hasn’t proven itself on your parts. Our engagement is built so you can stop — or continue — at each step, with the numbers in front of you.

// working together

Four steps, and you decide at each one

Step 1Low commitment · the recommended start

Line assessment

We look at the line itself: parts, defects, speed, lighting, constraints. Then we tell you whether vision or automation is the right answer, and roughly what it would take.

  • A written assessment of what’s feasible
  • A rough scope and indicative budget range
  • A clear recommendation — including “don’t” when that’s the honest answer
Step 2De-risk before you invest

Proof of concept

We prove the numbers on your parts and your defects before you commit to a full build. If it can’t hit the target, you find out early and cheaply — not after installation.

  • A working proof-of-concept on your samples
  • Measured results against agreed acceptance criteria
  • A fixed-price proposal for the full system
Step 3Fixed scope · fixed price

Turnkey project

Design, build, install and commission the system on-site, validated against the acceptance criteria until it holds the numbers under real production.

  • Installed, commissioned and validated system
  • Documentation and staff training
  • Handover with agreed acceptance sign-off
Step 4Ongoing · optional

Support & partnership

After go-live we keep the system performing — monitoring, tuning, retraining models as your parts change, and extending it as the line evolves.

  • Maintenance and response arrangement
  • Model retraining and re-tuning
  • A first call for the next improvement

// how pricing works

Fixed scope, fixed price, in writing

We don’t quote a full system blind. The line assessment gives you an indicative budget range early. The proof-of-concept keeps the first real spend small and proves the numbers. Only then do you get a fixed-price proposal for the build — with the scope, the acceptance criteria and the schedule written down.

No hourly meter running in the background, no surprise change orders for work that was always going to be needed. If scope genuinely changes, we agree it before, not after.

// delivery

How a build runs, step by step

01Assess

We look at the line itself: the parts, the defects, the throughput and the constraints. Then we tell you whether vision or automation is the right answer.

02Design

A concrete system architecture: sensors, lighting, compute, interfaces and acceptance criteria, with a fixed scope and price.

03Prove

A proof-of-concept on your samples before you commit to a full build. If it can’t hit the numbers, you find out early and cheaply.

04Build

Development in short iterations you can see, with the hardware and software validated against the acceptance criteria.

05Commission

Installation and tuning on-site, running against real production until it holds the numbers under real conditions.

06Support

Documentation, training for your staff, and a maintenance arrangement so the system keeps performing after handover.

// questions

Common questions

How do I know it will work on our parts?

You don’t have to take our word for it. Before any full build we run a proof-of-concept on your actual samples and measure it against acceptance criteria we agree up front. If it can’t hit the target, you know before you’ve committed to a system.

What does a project cost?

It depends on the line and the goal, so we don’t quote blind. The line assessment gives you an indicative range, and after a proof-of-concept you get a fixed-price proposal — no open-ended hourly surprises.

How long does it take?

A line assessment is days. A proof-of-concept is typically a few weeks. A turnkey system depends on scope and hardware lead times; a concrete schedule is part of the proposal.

Do we own the system and the data?

Yes. You own the deployed system and the production data it generates. We build on open, standard tools wherever possible so you are never locked to us to keep it running.

Is our data sent to the cloud?

Not by default. Inference runs on-premises, at the edge, next to the machine. If you want cloud reporting or dashboards, that’s a choice you make — not something baked in.

Can you work with our existing PLCs and MES?

Yes. Integration is the point. We connect to Siemens, Beckhoff and CODESYS control, and to your MES/ERP over OPC UA, MQTT and standard interfaces, rather than replacing what already works.

Do you work remotely or on-site?

Both. Development and much of the engineering happen remotely; assessment and commissioning happen on-site. We’re based in Poland and work across the DACH region and the wider EU, in English, German and Polish.

// step one

Book a line assessment

The smallest possible first step. Tell us about the line and you’ll get a straight read on what’s possible, and what it would take.