Controls migrations fail on planning, not on wiring. How we structure pre-builds, parallel installations and fallback points so a PLC retrofit fits inside a weekend instead of a shutdown.
The most expensive part of a controls retrofit is rarely the hardware. It is the production you do not run while the cabinet is open. A migration that needs a two-week shutdown will be postponed year after year; one that fits inside existing maintenance windows actually happens. So we plan retrofits backwards from the available downtime, not forwards from the engineering.
The first rule is that nothing gets engineered on site. The new control system is built, wired and tested in our workshop against a full EPLAN model of the existing installation, including the undocumented changes we find during the survey. Site time is for disconnecting, connecting and verifying — never for design decisions.
The survey deserves its own budget line. Twenty-year-old machines rarely match their drawings; terminals have been rerouted, spare cores pressed into service, safety circuits modified. We trace and document the as-built state before quoting the migration window, because every undiscovered deviation on site costs an hour you did not plan.
Where the machine allows it, we split the migration into stages with a working fallback at each one. Swap the operator interface first, run production on it for a week, then migrate the I/O, then the drives. Each stage has a defined point of no return, agreed with the plant, before which the old system can be restored in hours.
Simulation carries the rest of the risk. Before the window opens, the new PLC program has already run against a simulated model of the machine and passed a factory acceptance test with the customer present. What remains for site acceptance is confirming that reality matches the simulation — a checklist exercise, not an exploration.
A realistic weekend window for a mid-size machine looks like this: power down Friday evening, mechanical and electrical swap through Saturday, dry runs Saturday night, production trials Sunday, sign-off Sunday evening with the old panel still in the building. If a plan has no fallback and no margin, it is not a plan. It is a hope with a Gantt chart.
— GANI Engineering engineering team