PLC Replacement or Retrofit 2026 — Four Options for Ageing Control Systems

“We opened the control panel and the relay housings had turned brown.” “The CPU unit failed, and when we contacted the manufacturer we were told repair service had already ended.” These are familiar conversations in Japanese-owned factories across Thailand. The mechanical side of the equipment still runs well, but the control hardware inside reaches the end of its life first. At the same time, there is no budget to replace the entire machine. Deciding how and when to update or replace a PLC is a problem almost every plant eventually faces. This article sets out the yardsticks that separate life extension from full renewal, and organises the four realistic options in front of you.

Why control system obsolescence is becoming urgent now

When people hear “ageing equipment,” most picture wear in mechanical elements such as motors, bearings and cylinders. In practice, though, what stops a line first is often what sits inside the control panel. A worn mechanical part can usually be machined or rebuilt. A discontinued CPU unit cannot, because the replacement part no longer exists anywhere in the world.

Discontinuation and end of repair service are two different deadlines

The first thing to grasp about PLC lifespan is that manufacturers set more than one deadline. Control products typically carry at least these two.

• Discontinuation of production — the date new units stop shipping. After this date you can no longer buy that model number new.
• End of repair service — the last date the manufacturer will accept a failed unit for repair. After this date, genuine repair disappears as an option.

Plants that reassure themselves with “there is still stock available” are usually looking only at the first deadline. The operational risk actually arrives with the second. Around the discontinuation date you can still get by on market stock and in-house spares. Once repair service has ended and your spares run out, there is no longer any way to keep that machine running.

Roughly seven years separate the two deadlines

A concrete example helps. Mitsubishi Electric publishes the following schedule for parts of its MELSEC series.

The most important pattern in this table is that the gap between discontinuation and end of repair service is roughly seven years in every case. In other words, the moment you receive a discontinuation notice is the moment you should start making decisions about that machine’s control system, and a planning horizon of about seven years before genuine repair closes off is a realistic assumption.

This is not unique to Mitsubishi Electric. Omron maintains a dedicated PLC replacement portal listing discontinued products, recommended successors and migration guides. Every major PLC vendor now publishes end-of-life information and structured migration paths as a matter of course. Seen from the other side, this tells you something important. Vendors publish this material on the assumption that you will eventually migrate, and they have no intention of supporting previous generations forever.

Parts procurement is getting harder worldwide

Some pressures are entirely outside your control. As of 2026, engineers working in automation report that product model cycles have shortened compared with the past, and that sourcing components for previous-generation control systems has become harder or that lead times have stretched. This is not a figure backed by a specific statistic, but it is a trend enough practitioners describe in common that it is worth factoring into your planning.

For a factory in Thailand this cuts twice as deep. In Japan you can reasonably search the used market or find specialist repair shops. In Thailand the odds of locating the same model number second-hand domestically are lower, and importing from Japan or Singapore adds customs clearance and shipping days on top. The strategy of “start looking once it breaks” is becoming less viable with each passing year.

What happens if you leave an ageing control system alone

“It still runs, so we will deal with it when it breaks” looks like sensible cost control at first glance. For control systems specifically, however, that judgement tends to become expensive, and for structural reasons.

The damage from unplanned downtime becomes unpredictable

A planned update lets you pick a slow season, notify customers of the line stoppage in advance and arrange alternative production. An unplanned failure gives you none of that. Worse, an unplanned failure when parts have dried up creates a situation where you simply cannot restore the line until a component arrives, turning what should have been a few hours of downtime into several weeks.

When estimating the loss, the repair cost itself is rarely the largest number. Lost margin during the stoppage and the damage to customer confidence from a missed delivery usually dwarf it. In industries that require sequenced delivery, such as automotive parts, a single extended stoppage can affect the commercial terms of the relationship itself.

Security and compatibility problems accumulate

Previous-generation PLCs and their peripherals were never designed with network-borne attacks in mind. The moment you connect them to a production management system or MES to pull actual production data, the assumption of a closed network collapses. Programming tools that run only on obsolete operating systems are another recurring headache. It is far from unusual to find a plant where the ladder logic can only be edited on one surviving PC, and if that PC dies, nobody can modify the program at all.

Looking upward through the stack, older control hardware either does not support industrial Ethernet or supports it only in a limited way. The problem surfaces as soon as you try to make progress on factory data utilisation, with the generation of the control system acting as the constraint. For the wider picture of digitalising a plant, see Factory Automation in Thailand — The Full Picture of FA Deployment.

Maintenance becomes person-dependent and cannot be handed over

The most overlooked risk in control system obsolescence is the human one. For equipment installed twenty years ago, the design intent behind the ladder program may exist only in the head of the engineer who wrote it. Drawings never updated, no change history, no comments in the code. If that person resigns or returns to Japan, the line becomes a black box nobody understands.

In Thai plants, Japanese managers typically rotate every few years, so this dependency risk builds faster than it would at a domestic Japanese factory. A control system replacement is therefore not only a hardware refresh. It is also the opportunity to bring the program and drawings back in line with the current specification so that anyone can maintain the line. This secondary benefit is regularly left out of the cost-benefit calculation.

Yardsticks for the decision — extend or replace

So when should you stop extending and commit to replacement? Here are three yardsticks that let you decide with numbers rather than instinct.

Yardstick 1 — annual repair cost divided by acquisition value

A widely used rule of thumb in manufacturing equipment renewal is the ratio of annual repair spend to the original acquisition value of the asset. Once that ratio passes roughly 10 to 15 percent, it is generally treated as the signal to start evaluating replacement.

The logic is straightforward. Spending 15 percent of acquisition value on repairs every year means that in a little under seven years you will have spent the price of a new machine on keeping the old one alive. And repairs only restore the original performance. They do not increase throughput or reduce power consumption. Since the same money spent on new equipment buys performance improvement and energy savings at the same time, switching to renewal once the ratio climbs is the rational move.

The practical caution when using this yardstick is deciding what counts as repair cost. If you count only parts and outside contractor labour, the ratio comes out low. It only approaches reality once you include your own maintenance team’s hours, the margin lost to downtime and the extra cost of alternative production.

Yardstick 2 — the trend in MTBF

The second yardstick is MTBF, or mean time between failures. Dividing total operating hours by the number of failures tells you how many hours the machine runs, on average, before it stops.

MTBF is most useful as a trend rather than an absolute figure. If a machine’s MTBF was 2,000 hours three years ago and has fallen to 600 hours this year, that is a sign that many components are reaching end of life together. It indicates a shift from fixing individual faults one at a time to renewing the system as a whole. Conversely, if the absolute repair spend is high but MTBF is stable, you are probably looking at a design weakness in one specific area rather than general ageing, and addressing that one area may be enough.

Yardstick 3 — do not confuse statutory useful life with economic life

The statutory useful life used for tax depreciation and the period over which a machine remains economically sensible to run are two different things. Arguing that you should keep running a machine because it is fully depreciated misses the point, and so does arguing that you must replace it because it has passed its statutory life.

The recommended practical approach is to lay the projected cumulative repair spend over the next five to ten years alongside the acquisition and installation cost of new equipment. Framed that way, you can debate which is cheaper on common ground, regardless of how far depreciation has run.

When extending the life is the right call

Not every ageing machine should be replaced. Life extension is a rational choice when the following conditions hold together.

• Annual maintenance spend is under 10 percent of what a replacement would cost
• Replacement parts remain obtainable, either on the market or from internal stock
• No investment plan exists this fiscal year that could absorb a large renewal cost
• Production of the product this machine makes is scheduled to end within a few years

The important qualifier is that life extension is rational only when these conditions overlap, not when one of them happens to hold. The pattern to avoid most is continuing to extend on the strength of the parts-availability condition alone, then realising that the end of repair service passed several years ago.

Comparing the four options

In practice, responses to control system obsolescence fall into roughly four options. Laying all four side by side is the starting point for the decision.

Option 1 — status quo plus spare parts stocking

This means changing nothing and extending service life by securing spares in a planned way. When a discontinuation notice arrives you buy the quantity you expect to need, and when something fails you swap in your own spare. Additional investment is the smallest of the four options, and no line stoppage is required.

The weakness is that this only buys time. Once the spares are used up, that is the end. The stored spares themselves also degrade with age, for example through electrolytic capacitor deterioration, simply from sitting on the shelf. This approach also defers the ageing problem without solving anything on the connectivity or security side.

Option 2 — retrofit while keeping the existing control system

Here the existing panel and PLC stay in place and the functions you need are added externally. Retrofit in this sense covers a range, from conservative work that replaces only the most degraded components inside the panel, through to tapping the existing PLC’s output signals and collecting data via an external gateway.

Note that retrofits whose main purpose is to add sensors and gateways to capture operating data, in other words IoT enablement, follow a different cost structure and a different approach. That topic is covered in detail in IoT Retrofit for Legacy Equipment — The Five Cost Layers and How to Proceed in a Thai Plant. The retrofit discussed in this article is strictly about extending the life of the control system itself, so the objectives differ. If your goal is data collection, start from that article instead. If your goal is extending the life of the panel, use the framework in this article.

Option 3 — replacing the PLC and control panel

This option keeps the mechanical equipment as it is and swaps the contents of the panel, including the PLC, for a current generation. The ladder program is migrated to the new platform and wiring is re-run where necessary. The cost is lower than replacing the whole machine, yet the control system’s life resets to that of new hardware, and connectivity to host systems and security both rise to current standards.

The practical questions when choosing this option are how the program will be migrated and how much of the existing I/O wiring can be reused. Some vendors offer conversion adapters matched to the terminal layout of the older model, which can allow the existing wiring to be carried over directly. Whether that is available changes both schedule and cost significantly, so it is worth confirming before you request quotations. On where to have the new panel built, the trade-offs between building in Japan and building locally in Thailand are set out in Control Panel Design and Manufacturing 2026 — Build in Japan or Build in Thailand.

Option 4 — replacing the equipment outright

This means replacing the whole machine, mechanical elements included. Not only the control system but also mechanical accuracy return to as-new condition, and you can gain throughput, energy efficiency and conformity with current safety standards at the same time.

Cost and downtime are the highest of the four. This option is justified when mechanical wear has reached its limit alongside the control system, or when you need to produce something the current machine cannot make. The decision criteria from a whole-lifecycle perspective are covered in Legacy Equipment Replacement 2026 — Four Deadlines Before the Parts Vanish, which is worth reading if you also need to judge the mechanical side.

Comparison table

The table below places the four options side by side in terms of cost, downtime and the situations they suit. Amounts vary enormously with the scale and I/O count of the equipment, so read these as relative tendencies rather than absolute figures.

Each option also carries its own risks, and choosing purely on the lowest price tends to cost more later. The points most often overlooked are set out below, in the same order as the four options.

Considerations specific to Thailand and ASEAN

Everything so far applies equally to a plant in Japan. For manufacturers operating in Thailand, a few local factors are added to the mix.

Factor in BOI technology upgrade incentives

In 2026 the Thailand Board of Investment (BOI) is running its Smart and Sustainable Industry framework to encourage investment in smart factories, AI-assisted production and automation. What deserves attention is that the scheme is not limited to new factory construction. There is a provision (Activity 10.1) under which already-operating plants can apply for incentives when carrying out technology upgrades.

For qualifying technology upgrade investment, additional years of corporate income tax (CIT) exemption may be granted, described in some sources as up to around three years depending on the project. There are also import duty exemption schemes covering AI and automation-related equipment.

That said, the specific conditions such as exemption rates and caps are described differently across sources, and application requirements are updated from year to year. Whether a control system update qualifies under this framework depends on the breakdown of the investment and the nature of the technology involved, so always confirm the details of the incentives and the current application requirements on the official BOI website and by contacting BOI directly. What appears here is only an indication that such schemes exist.

The practical point to keep in mind is that the availability of incentives can reverse the decision itself. Even where the annual repair cost ratio alone would suggest extending the life, replacement may come out ahead if incentives reduce the effective investment burden. It is well worth checking applicability with whoever owns your investment planning before you finalise the decision.

Rising labour costs and the shortage of skilled staff

Labour costs in Thailand continue to rise while skilled workers remain in short supply, and multiple industry reports point to this as an accelerant for automation investment, particularly in automotive and electronic components.

That context feeds directly into control system decisions. Lines built around previous-generation PLCs often rely on an experienced operator’s intuition for changeovers and fault recovery, and the cost of retaining people with that intuition climbs every year. Moving the control system to a current generation so that changeovers become recipe-managed and fault guidance appears on screen is not merely a hardware refresh. It is an investment that lowers the skill requirement for the people you need. That effect never shows up in a repair cost comparison, but over a few years it becomes too large to ignore.

Mixed vendor environments and choosing a local system integrator

Japanese-owned plants in Thailand commonly run a mixture of equipment brought over from Japan, equipment sourced locally and equipment relocated from another site of the parent group. The result is several PLC brands sitting side by side in one factory, each needing its own development environment and its own spare parts inventory.

A control system update is one of the few opportunities to tidy that mixture up. There is no need to standardise everything at once, but if each update follows a stated policy of which vendor and which family will be the plant standard from now on, you can compress the number of spare part types and the maintenance learning curve over several years. Keep choosing whatever is cheapest at the moment and the mixture only gets worse.

When selecting who does the work, three questions serve as realistic screening criteria. Do they have hands-on experience with the PLC brand in question, can they read and migrate your existing ladder logic, and can they provide after-sales support from within Thailand? Why quotations diverge so widely and which boundaries to confirm are covered in How to Choose a Robot System Integrator 2026 — Five Boundaries That Make Quotes Differ by Two or Three Times, which is useful reading at the stage of comparing suppliers.

How to proceed — from assessment to phased execution

Once the criteria are settled, the next question is sequencing. Attempting a control system update all at once makes neither the budget nor the downtime realistic, so a phased approach is the norm.

Phase 1 — inventory and assessment of the current state

Start by listing every PLC and control panel in the plant at model number level. The information to gather is vendor, model number, year installed, discontinuation and repair service status, number of spares held, the state of the program and drawings, and the fault history over the past few years.

At most plants this inventory turns up surprises. Zero stock of a model number everyone assumed was covered, an end of repair service date that passed years ago, uncertainty about which version of the ladder logic is current. Half the value of an assessment lies exactly there, in finding out what you did not know.

Phase 2 — setting priorities

Sort the listed equipment on two axes, risk and criticality. Risk is judged from the time remaining until end of repair service and the number of spares on hand. Criticality is judged from the production impact if that machine stops.

Anything high on both, meaning repair service has already ended, no spares remain and a stoppage halts the entire line, is the top priority. Equipment that is high risk but where production can be shifted elsewhere can wait. Skip this step and work through the list oldest first, and you risk exhausting the budget while the most dangerous machine is still untouched.

Phase 3 — a focused pilot

Run the highest-priority machine, or one line, ahead of the rest. What you want out of it is the actual hours and cost incurred, the gap between estimate and outcome, and the kinds of problems that arose during migration. Having that record transforms both the accuracy of subsequent quotations and your ability to make the internal case.

When choosing the pilot, note that the most dangerous machine and the most difficult machine are not always the same one. Picking something extremely difficult first risks stalling partway and taking the whole plan down with it. A machine that is high risk and moderate in difficulty is the realistic choice.

Phase 4 — rollout and standardisation

Use what the pilot taught you to work through the remaining equipment. This is the stage at which to fix the platform standard, the coding conventions for programs, how drawings are managed and which spare parts are shared across machines. Optimise each update in isolation and all you create is a new mixture.

In a Thai plant, we strongly recommend preparing this standardisation documentation in Thai as well. Procedures available only in Japanese and English rarely function as day-to-day reference material for Thai maintenance staff, and that is a common reason why hard-won maintainability erodes again within a few years.

Frequently asked questions

How much does a PLC update typically cost?

There is no single figure. Cost varies widely with I/O point count, whether existing wiring can be reused, the size of the ladder program and the state of the drawings, and how much downtime is available.

The realistic approach is not to hunt for a market rate but to compare, for your own equipment, the projected cumulative repair spend over the next five to ten years against the acquisition and installation cost of a replacement. That framing lets you decide without worrying whether someone else’s numbers apply to you. Checking whether annual repair spend has passed 10 to 15 percent of acquisition value gives you a further indication of whether you have reached the point of evaluating replacement.

What is the difference between a retrofit and a PLC replacement?

The objective and the scope of the work. A retrofit is a life extension measure that adds functionality or replaces degraded components while leaving the existing control system in place. It keeps cost and downtime low, but it does not extend the life of the control hardware itself. A PLC replacement moves the control side to a current generation. Cost and downtime are higher, but the control system’s life resets and connectivity to host systems and security come up to current standards.

Note also that the word retrofit is often used to mean adding IoT sensors to capture data, which is a different objective from extending the life of a panel. It is an easy place for internal discussions and vendor conversations to drift apart, so defining what the retrofit is meant to achieve at the outset keeps everyone aligned.

What happens if we leave an ageing PLC as it is?

The most direct risk is a failure after repair service has ended, where parts cannot be obtained and there is no clear path to recovery. A fault that would normally mean a few hours of downtime can stretch to several weeks while a component is located and imported.

On top of that, three problems advance in parallel. Losing the last PC that runs the old development environment means you can no longer even edit the program. Connectivity to host systems and any meaningful security measures remain out of reach. And once the people who understood the design intent have gone, maintenance can no longer be handed over. This combination is precisely why waiting for something to break no longer works as a strategy.

What does a PLC update project in Thailand usually look like?

A typical sequence runs from inventory and assessment of the current state, to defining the target equipment and scope, investigating the existing program and drawings, selecting the platform and deciding the migration approach, designing and building the control panel, installation and cutover within a secured downtime window, and finally commissioning and witnessed acceptance.

The Thailand-specific considerations layered on top are whether to build the panel in Japan and import it or build it locally, confirming eligibility for BOI incentives, and verifying that post-update maintenance can be completed within Thailand. That last point especially needs checking while you are still selecting a supplier, or you can end up with an updated system that requires flying someone in from Japan whenever something goes wrong.

Should we update everything in one go?

As a rule we recommend a phased approach. Updating everything at once not only concentrates the budget peak but usually requires a long line stoppage, with an impact on the production plan that exceeds what most operations can absorb.

There is an exception. Where several machines are linked on the same network and moving only some of them to a new generation would break the communication specification, executing that linked group together generally works out cheaper. Think of the phasing unit not as a number of machines but as the smallest unit that can be functionally separated.

Summary

Control system obsolescence can halt a factory before mechanical wear ever does. The key points are as follows.

• Manufacturers set two deadlines, discontinuation of production and end of repair service, and the operational risk arrives with the second. In the Mitsubishi Electric examples, roughly seven years separate the two
• The yardsticks for the decision are whether annual repair cost has passed 10 to 15 percent of acquisition value, whether MTBF is trending downward, and which is larger, cumulative repair spend over the next five to ten years or the cost of renewal
• The options are status quo plus spares, retrofit, PLC and panel replacement, and full equipment replacement. Compare them on their specific risks, not only on cost and downtime
• In Thailand, three local factors shape the decision — BOI technology upgrade incentives, rising labour costs alongside a shortage of skilled staff, and mixed vendor environments
• Proceed in phases, in the order of inventory and assessment, prioritisation, a focused pilot, then rollout and standardisation

The outcome to avoid above all is deferring the decision itself while time passes. Roughly seven years from a discontinuation notice to the end of repair service is not a long window. Starting with just the inventory and assessment gives you the room to get renewal onto a budget plan.

If you are unsure where your own panels and PLCs stand, or whether extension or replacement is the right call, that uncertainty is a perfectly good starting point for a conversation. TOMAS TECH designs, builds and maintains FA systems in Thailand, and we are happy to begin with an inventory and current-state assessment of your existing equipment. Get in touch through our contact page and we can start by putting your current situation in order.

References

• Mitsubishi Electric MELSEC — Discontinued Products and End of Repair Service
• Omron PLC Replacement Portal
• Judging Equipment Renewal for Ageing Assets in Manufacturing – Yachiyo Solutions
• Challenges and Countermeasures for Equipment Ageing – Yachiyo Solutions
• Depreciation and Useful Life of Machine Tools – Monozukuri Base by Yamazen
• The 2026 Challenge – Sourcing Obsolete PLC Parts and Maintaining Legacy Automation Systems
• Thailand BOI Incentives for Manufacturing – Pertama Partners
• Thailand’s BOI in 2026 – From Investment Incentives to Accelerated Project Delivery
• Thailand Board of Investment (BOI) Official Website