

The cheapest automation project is the one that does not replace anything. A line bought second-hand usually has years of mechanical life left and a control system nobody wants to touch. The instinct is to rip out the controller. The better first move is almost always to read the machine before you rewrite it.
This is the companion to moving production to Poland: what to do once the hall is yours and the line in it was somebody else's.
Before anyone quotes you anything, establish three things:
Here is the counter-intuitive part. Before you change what the line does, measure what it does. A read-only data layer — taken from existing contacts, retrofitted sensors and drive parameters — is cheap, changes no logic, carries almost no risk of stopping production, and tells you where the losses actually are.
That matters because retrofit budgets are routinely spent on the wrong station. The bottleneck people assume and the bottleneck the data shows are frequently different, and an intervention on the wrong one buys nothing. Calculating OEE per asset is the standard way to find out; connecting older machines and CNC tools is how the signal gets out of a line that predates the idea.
With a month of real data, the decision gets simple. Replace the controller when it is unsupported, when the logic has to change for a new product, or when the program is unrecoverable. Otherwise, fix the station the data indicted and leave the rest alone.
And know what you are buying at each level before you commit — monitoring, MES and ERP solve different problems and get confused with each other constantly: MES vs SCADA vs ERP. The cost question is answered honestly in what a production monitoring system costs.
A retrofit can change a machine's compliance position. If you are modifying safety functions or substantially changing what a machine does, that is a question with a legal answer, not an engineering one — EU Machinery Regulation 2023/1230 is where it starts. Raise it before the project, not during the audit.
We work on this with a manufacturing partner and a Polish contract. Tell us what the line is and we will tell you what we would measure first.
Not as a first move. Replace it when the platform is unsupported, the logic must change for a new product, or the program source is unrecoverable. Otherwise measure the line first — the bottleneck you assume and the one the data shows are often different.
What the controller is and whether it is still supported, and whether the program source files exist — in writing, before payment. A controller you cannot open is a controller you can only replace.
It can. Modifying safety functions or substantially changing what a machine does is a question with a legal answer rather than an engineering one — EU Machinery Regulation 2023/1230 is where it starts. Raise it before the project, not during an audit.
Newsletter

By clicking the submit button, you agree to the
rules for processing personal data.