When to modernize a control system — and when to wait
Obsolescence risk, spare-parts reality and the true cost of an unplanned outage — a framework for timing a brownfield retrofit, from the engineers who run the cutovers.
Published 15 Sep 2026 · 7 min read
"Should we modernize now?" is the wrong first question. The right one is: what actually breaks if we don't — and when? Most plants we work with wait too long on systems that are genuinely at risk, and rush systems that still have years of safe life left, because the decision gets made on age instead of on the three things that actually matter: parts availability, vendor support status, and how much unplanned downtime the plant can absorb.
This is what the industry calls a brownfield project — retrofitting or revamping the control system on a line that's already producing, as opposed to a greenfield build on a clean sheet. The engineering discipline is different: every design decision has to account for what's already installed, what can keep running during the work, and how the old and new systems hand over to each other without a full shutdown.
The real risk isn't the PLC. It's the spares bin.
A control system rarely fails all at once. It fails one module, one power supply, one HMI panel at a time — and the question that decides whether that's a five-minute swap or a five-day production stop is whether a replacement part exists anywhere in the world. Once a platform hits end-of-sale, the clock on end-of-support usually isn't far behind, and after that, every failure becomes a scramble: refurbished parts of unknown history, eBay sourcing, or a full emergency migration done under the worst possible conditions — unplanned, unbudgeted, and with production stopped.
That's the framework we actually use with customers: modernize on a schedule you control, before the schedule gets forced on you by a failed module and no spare.
Four questions before you commit
- Is the platform still sold, and for how long? Vendor end-of-sale and end-of-support dates are public — check them before you check your budget.
- Can you still get spares within your acceptable downtime window? Same-day availability from a distributor is a different risk profile than a six-week special order.
- Does the current system block something the business actually needs? A new recipe, a new reporting requirement, an integration a customer is asking for — obsolescence is only half the case; the other half is opportunity cost.
- What's the real cost of an unplanned failure versus a planned migration? Planned migrations are scheduled around production; unplanned failures aren't. Price both scenarios honestly, including lost production, not just parts and labor.
When the answer is "wait"
Modernizing too early wastes capital and disrupts a system that's working. We tell customers to wait when the current platform still has clear vendor support, spares are readily available, the system isn't blocking any near-term business need, and there's a bigger plant change coming in the next 12–24 months that a migration should be bundled with rather than run twice. A control system with years of supported life left doesn't need a new brain; it needs good maintenance discipline and a documented spares inventory.
When the answer is "now"
The signal we treat as non-negotiable is a vendor end-of-support date inside your planning horizon combined with single-source or no-longer-manufactured spares. At that point, every month of delay is a month closer to an unplanned outage with no fallback — and unplanned outages cost far more than the migration itself, in lost production, expedited freight, and overtime labor, before you've even fixed the underlying problem.
A migration doesn't have to mean a shutdown
The part that scares plant managers most isn't the decision to modernize — it's the assumption that it means stopping the line. It usually doesn't. Staged cutovers, running the new and old control layers in parallel during commissioning, and moving one process cell or one batch stage at a time let us validate the new system against real production before the old one is retired, with a fallback path at every step. The projects that go badly are the ones planned as a single big-bang cutover under time pressure — which is exactly the scenario a planned, early migration is designed to avoid.
If you're not sure which side of that line your system is on, that's usually a half-day conversation, not a guess: a spares and support-status review against your actual production risk tolerance tells you whether you have years of runway or months.