Equipment Modification Cost and Decisions 2026 – 4 Gates

You asked three vendors to quote the same equipment modification, and the highest came back at nearly double the lowest. Plant managers in Thailand hear this story constantly. The specification was identical. The jig to be added was identical. The screw-tightening unit was identical. And still the quotes diverged. The cause is almost never the amount of work. It is what your own organisation still holds on file about the machine in question. This article breaks down the decision between modifying existing equipment and building new, using decision gates and a layer-by-layer cost model you can take into an internal review meeting and defend with numbers.

Why equipment modification quotes diverge – surviving information, not workload

The same scope, two very different prices

When a plant asks us about an equipment modification, what usually arrives first is a list of what they want added. Add one set of workholding jigs so the line can run a new model. Add one screw-tightening axis. Keep the cycle time where it is. As a statement of requirements, that is sufficient. What surprises people is that the quotes generated from those requirements swing more with the volume of surviving documentation than with the engineering capability of the companies quoting.

The reason is straightforward. Any job that touches an existing machine has to establish what the machine currently is before a single part can be designed. On a new build, the machine and its control system are whatever the builder decides, so that establishing work is absorbed into design. Retrofitting existing machinery is different. You are placing today’s requirement on top of decisions somebody else made more than a decade ago. If those decisions were never written down, the only remaining source is the machine itself, and someone has to read it back out. That reading effort is exactly what shows up as the spread between quotes.

The cost of re-discovery never appears in a drawing or a spec sheet

The awkward part is that re-discovery produces no visible deliverable. From the plant floor, the money bought one jig set and one extra axis. Redrawing the wiring schematic from the panel, reverse-reading the PLC program, dumping the servo parameters and inferring why each value was set that way – none of that changes the appearance of the running machine by a millimetre.

So the internal capital request tends to come back with the comment that this is an expensive way to buy one jig. The mental model a buyer needs here is that every equipment modification quote contains two different kinds of money mixed together. There is the cost of making things, and there is the cost of finding out what is already there. Ask for those two to be presented separately and you can judge whether the price is reasonable. A single lump-sum figure cannot be compared against anything.

Before you assess the vendor, assess your own filing cabinet

When a modification enquiry reaches us, the first thing we ask to see is not the machine. It is the machine’s file. Mechanical drawings and electrical schematics. PLC source code and HMI screen data. The servo parameter table. And the record of what has been changed since handover. How much of that set still exists determines both the difficulty and the price of the modification before any vendor gets involved.

Put differently, half of the modify-or-replace decision is settled inside your own building, before you call anyone. Opening the cabinet and checking what is there costs nothing and can be done this week. In the worked model later in this article, the identical modification comes out at 1,400,000 THB when the records are complete and 2,280,000 THB when they are missing. The difference is 880,000 THB.

Why equipment modification enquiries are rising in Thailand in 2026

What the recent production statistics actually say

According to Xinhua on 27 July 2026, Supakit Boonsiri, Director-General of the Office of Industrial Economics under Thailand’s Ministry of Industry, announced that the Manufacturing Production Index for June 2026 fell 3.10% year on year. For the second quarter of 2026 the decline was 1.79%. Average capacity utilization over the period was reported at 57.47%. On the June figure, Business Recorder noted that the result came in below market expectations, as a Reuters poll had forecast a 0.2% increase.

Read literally, that looks like a signal to stop investing. It is not. It needs translating into capital-planning terms, and the translation runs in two directions at once.

Capacity utilization of 57.47% cuts both ways

The first reading is the obvious one. At 57.47% utilization, a proposal to add capacity is hard to get approved. When idle capacity already exists, adding another machine to increase output makes the payback argument difficult to write. Volume-driven new-build projects stall in this environment.

The second reading pulls the other way. Slack capacity means the line can be stopped. A plant running flat out against a full order book may need months of negotiation just to secure three consecutive days of downtime. When downtime cannot be obtained, modification becomes impossible in practice, and the decision drifts toward a new build precisely because a new build can be assembled off-line and only needs a short installation window. A period of lower utilization removes that constraint.

Which is why the enquiries moving are about variants, not capacity

The enquiries actually increasing are not about producing more units. They are about making existing equipment run something it could not run before. A new model has been won. The customer has issued a specification change. Another part number needs to go down the same line. Requirements of that type arise whether or not volume is growing. In fact, the less predictable the volume, the more a plant leans toward making one existing machine handle multiple variants rather than buying a second dedicated machine.

Most of the production line modification work under consideration in Japanese-owned plants in Thailand right now sits in that context. And variant-capability work reaches deeper into the existing machine than capacity work does. It touches control logic, changeover procedure, jig datums, guard openings. In other words, it is exactly the category of modification where surviving documentation matters most.

The four gates that separate modification from new build

Run the gates in order, not in parallel

Whether an equipment modification is viable can be judged in a fixed sequence rather than by instinct. The four gates are information, structure, safety and downtime. They are not equally weighted alternatives to be scored side by side. They cascade. Without information you cannot measure structural headroom, and without both of those you cannot assess safety properly.

One red gate does not kill the project

If one of the four gates comes back unfavourable, the modification is not dead. What matters is knowing in advance which part of the cost that unfavourable gate will land on. Missing information lands on the survey layer. Insufficient structural headroom lands on the mechanical layer. A required safety re-evaluation lands on the safety and documentation layer. A longer stoppage does not land on the quote at all – it lands on lost production.

The inverse is the useful rule. When all four gates are unfavourable, do not decide on the modification in isolation. Quote the new build at the same time and compare. In the model below, the gap between a modification with missing records at 2,856,000 THB all-in and a new build at 3,492,000 THB all-in narrows to 636,000 THB. That difference does not buy you remaining mechanical life.

Gate 1 – Information, meaning drawings, PLC program, parameters and change history

What “complete” actually means for each of the four

This gate is not a check on whether files exist. It is a check on whether the files agree with the machine.

For drawings, the mechanical set needs assembly drawings and key part drawings that reflect the state of the machine after the last time anyone touched it, and the electrical schematic needs to match the actual wiring and terminal numbering inside the panel. For the PLC, source code has to exist rather than a compiled binary, and the comments have to have survived. For servo parameters, both the values and the reasoning behind those values need to be recoverable. For change history, it must be possible to trace who changed what, when and for what reason.

Of the four, the one most often missing on a real shop floor is change history. Drawings and programs from the original handover are still in the cabinet, but the small corrections made on site over more than a decade were never recorded. That state is worse than having nothing. Because documentation exists, the quote is written on the assumption that it matches, and the discrepancy is only discovered after the machine is opened up and the clock is running.

For a PLC program, readable matters more than present

With PLC programs, the data frequently exists yet is unusable in practice. Ladder comments were written in Japanese, and after years of maintenance by Thai engineers the comment language is mixed. There is no device allocation table, so nothing in the program explains what M100 represents. Only a compiled file survives, and nobody knows which version of the development environment produced it.

Modifying from that starting point takes more engineering hours than writing new code. To guarantee that existing behaviour is not broken, someone has to read the entire running logic first. How much of that reading to outsource, and where to draw the boundary, is covered in our article on outsourcing PLC program development.

Information is won or lost at installation, not at modification

The root cause of missing information is almost never the modification project. It is the original installation. Commissioning teams are under delivery pressure, and the project tends to be declared complete the moment the machine runs. Unless receipt of source code, a parameter backup and a defined change-history format are written into the acceptance conditions, the omission comes back as hard cash a decade later. We deal with that handover set in our article on equipment startup support.

For machines where the information is already gone, it is faster to place them inside a replacement plan rather than treating each modification as an isolated decision. The framework for that is set out in legacy equipment replacement.

Gate 2 – Structure, meaning rigidity, space, axes and power capacity

Four kinds of headroom to check

The structural gate asks whether what you want to add will physically fit and function. Four things need checking – rigidity, space, axes and power capacity.

Rigidity is whether the existing base frame and structure can carry both the mass of the added jig or unit and the reaction forces it generates in motion. When the addition is something like a screw-tightening unit, static mass alone is not a sufficient test. If the original frame was designed close to its assumed load at the time of installation, positioning accuracy will degrade the moment the new unit is bolted on.

Space is not about whether the new hardware fits in the envelope. It is about whether a maintenance technician’s hands and tools still fit. A modification that makes it impossible to get a tool onto an existing wear part will cause problems every time that part needs changing.

Axis count is whether the existing servo amplifiers and controller can accept another axis. Adding a fourth axis to a machine built around three servo axes runs into the controller’s axis limit, the number of stations the network supports, and whatever scan time margin remains.

Power capacity covers the panel’s main breaker rating and the thermal condition inside the enclosure. More devices mean more consumption and more heat rejected inside the panel. Even where physical space is free in the existing enclosure, the thermal budget may not close.

When headroom is gone, cost jumps to the mechanical and electrical layers

If any of the four falls short, the nature of the job changes from adding a jig to rebuilding a frame or replacing a panel. At that point the cost structure of the equipment modification starts to resemble a new build. Where panel replacement becomes necessary, the panel alone costs roughly what the control panel for a whole new machine would. The criteria for deciding whether an existing enclosure can be modified or should be replaced are covered in control panel design and manufacturing in Thailand.

Judge structure on the machine, not on the drawing

Some structural headroom can be calculated from drawings and some can only be measured on the machine. This is especially true of equipment that has been relocated. When a machine is moved from a plant in Japan to Thailand, it is disassembled for shipping and rebuilt on site. The foundation specification and the levelling accuracy of the base frame after that rebuild are not necessarily what they were in Japan. It genuinely happens that a drawing shows adequate headroom while the as-installed condition prevents the added hardware from holding tolerance.

For that reason, this gate requires at least one physical survey of the machine. That survey cost sits inside the survey layer of the five-layer breakdown below.

Gate 3 – Safety, because a modification can create a substantially new machine

The framework that bites when you ship to Europe or follow a European parent’s standard

On machinery safety, Regulation (EU) 2023/1230, dated 14 June 2023, applies from 20 January 2027 and replaces Directive 2006/42/EC. The Regulation carries the concept that a party who makes a substantial modification to a machine may be regarded as the manufacturer and may therefore take on obligations for conformity assessment, technical documentation and CE marking.

The scope needs stating precisely. This is an EU regulation. It is not legislation that applies directly to a factory in Thailand or in Japan. Where it does bite is when the modified machine, or product made on that machine, is shipped to Europe, and when a group with a European head office applies the thinking of the Machinery Directive or Machinery Regulation as an internal standard. Japanese-owned plants in Thailand with European customers, or sitting under a European parent, do sometimes find this framework handed down as a corporate requirement. Whether your site is inside that perimeter is something to confirm before the modification specification is frozen, not after.

Under the Japanese framework, changing equipment is a trigger for risk assessment

Risk assessment under Article 28-2 of Japan’s Industrial Safety and Health Act includes the point at which equipment is changed among its trigger events. It is, however, a duty of effort rather than a legal obligation, with certain areas such as chemical substances having been made mandatory. It is worth being careful not to misrepresent this in internal documents. A duty of effort is not the same thing as an obligation, and it is also not the same thing as something you can skip.

Separately, notification of machinery hazard information by a business operator that transfers or leases machinery is likewise treated as a duty of effort, under Article 24-13 of the Ordinance on Industrial Safety and Health. That concept becomes relevant when equipment is moved between group companies, or when a modified machine is transferred to another site.

After a modification, the information for use no longer matches the machine

Japan’s Comprehensive Safety Standards Guidelines for Machinery, revised by Notification No. 0731001 of 31 July 2007, provide that a business operator having workers use machinery should confirm the information for use supplied by the manufacturer, carry out the assessment under Article 28-2 on that basis, establish procedures for communicating residual risk to workers, and provide training.

This also needs reading carefully. The guidelines are centred on the safe use of machinery as supplied, and we have not been able to confirm a provision that directly governs modification carried out by the user. So it is better not to write in an internal document that the guidelines regulate modification. The sound practical position is this. Once you modify a machine, the information for use received from the manufacturer – the operating manual, the residual risk list, the content of the warning labels – no longer matches the machine in front of the operator. Training and procedures cannot be built on information that does not match, so the party performing the modification has to reconstruct it.

Rebuilding safety costs more in paper than in steel

Safety measures bring guarding and light curtains to mind, but in a cost breakdown the documentation side carries the greater weight. Repeating the risk assessment. Updating the residual risk list. Revising the operating manual. Revising the work standards. And delivering operator training in Thai. That whole package sits in the safety and documentation layer of the five-layer model. In the worked comparison, it is budgeted at 180,000 THB for the modification with complete records and 220,000 THB for the modification without them.

Gate 4 – Downtime, because days off-line move the total more than the quote does

Fix the cost of one day of downtime before anything else

The item most often missing from a modification capital request is the monetary value of the days the line will be down. The model in this article assumes eight hours a day and gross profit of 12,000 THB per hour, giving lost profit of 96,000 THB per day of line stoppage. That number differs from plant to plant, so substitute your own. What matters is that it is settled before the modification study begins.

Without a value per day of downtime, the modification and new-build options get compared on quoted cost alone. On quoted cost, the modification is obviously cheaper. But the modification takes the line down for longer. A new build that only needs installing takes the line down for a very different number of days than a modification that requires dismantling and reworking a running machine.

Missing information also lengthens the stoppage

Downtime is not driven by the volume of work. It is driven by how much can be fixed in advance. With drawings and program in hand, the post-modification logic can be written and simulated in the office before the line stops. What remains inside the window is mechanical fitting, wiring and functional checks.

Without records, none of that pre-work is possible. The wiring gets traced, the program gets read and the parameters get dumped with the machine already stopped. That difference is why the model uses three days for the modification with records and six days for the one without. Converted to lost production, that is 288,000 THB against 576,000 THB, a difference of 288,000 THB.

Whether the stop can be split changes the answer

Even a plant that cannot release a single long window can sometimes split the work. Mechanical fitting over one weekend and wiring over the next keeps the impact on output small. Splitting does add cost, though, because each stop carries its own shutdown and restart effort. That is a decision to spend more money in order to lose less production, and it cannot be evaluated at all unless the value of one day of downtime has already been agreed.

Read equipment modification cost in five layers

The five layers, plus installation and commissioning

A lump-sum modification quote cannot be compared and cannot be negotiated. We recommend requiring quotes to be broken into the following five layers, with installation and commissioning added on top since it arises in every option.

• Survey layer – site survey, physical measurement, redrawing schematics, program analysis, parameter extraction, specification freeze
• Mechanical layer – design and manufacture of jigs, units, frames and transfer sections, plus rework of existing mechanisms
• Electrical and control layer – panel modification or replacement, wiring, added sensors and actuators, network
• Software layer – PLC program modification, HMI screen modification, servo parameter tuning
• Safety and documentation layer – repeat risk assessment, added guarding, revision of operating manual and work standards, training
• Installation and commissioning, which sits outside the five layers because it arises in every option – site installation, adjustment, witnessed testing, handover

Missing information does not hit every layer equally

The reason for splitting into these layers is that the penalty for missing documentation is not distributed evenly. The mechanical layer builds the same hardware either way, so the increase is limited to extra measurement and fitting to the actual machine. The survey and software layers, by contrast, convert missing information directly into hours.

In the comparison below, of the 880,000 THB cost difference between the option with records and the option without, the survey and software layers alone account for 600,000 THB. That is 520,000 minus 120,000, or 400,000, on the survey layer, and 420,000 minus 220,000, or 200,000, on the software layer. Most of the gap is not the cost of making anything new. It is the cost of re-establishing what already exists.

The electrical and software layers move most with how you package the work

On the electrical and control layer, the order of magnitude changes depending on whether the existing panel can be modified or a new panel has to be built and swapped in. Where the existing enclosure is old, has no free space and no longer matches its drawings, building new is sometimes both cheaper and faster than modifying. The criteria are covered in control panel design and manufacturing in Thailand.

On the software layer, price moves with who performs the rework of the existing program. Returning to the original machine builder minimises the reading effort because the design intent is already there. If that builder is in Japan, travel cost and response time become the constraint. Switching to a company that can support the machine locally means the reading effort has to be explicitly priced into the quote rather than assumed away.

A worked comparison – modification with records, modification without records, and a new build

The model case

Fixing the conditions for comparison. One semi-automatic assembly machine in a Japanese-owned plant in Chonburi province, Thailand, relocated from Japan in 2012. Control consists of one PLC, one 10-inch HMI touch panel and three servo axes. The purpose of the modification is to add one set of workholding jigs and one screw-tightening unit to support a newly added model, while holding cycle time unchanged.

Lost profit for one day of line stoppage is 96,000 THB, as established above.

Option 1 – modification with drawings and program available

Downtime is 3 days, lost production is 288,000 THB, and the all-in figure is 1,688,000 THB.

Option 2 – modification with drawings and program missing

Downtime is 6 days, lost production is 576,000 THB, and the all-in figure is 2,856,000 THB.

Option 3 – new build of an equivalent semi-automatic machine

Downtime is 2 days, because the machine is built off-line and only installation happens on site. Lost production is 192,000 THB, and the all-in figure is 3,492,000 THB. How a machine like this is specified and built from scratch is covered in special purpose machine design and manufacture.

The ratios when only quoted cost is compared

On quoted cost alone, Option 1 is 42.4% of Option 3 and Option 2 is 69.1% of Option 3. The gap between Option 1 and Option 3 is decisive – where the records are complete, modification wins outright. Option 2, on the other hand, is close enough to the new build that the argument “modify because it is cheaper” stops holding up on its own.

Where the gap closes once downtime losses are included

The all-in ranking

Adding lost production to quoted cost gives the following.

The all-in gap between Option 2 and Option 3 is 636,000 THB. It is narrower than the comparison on quoted cost, because downtime runs the opposite way to price – 6 days for the modification against 2 days for the new build.

636,000 THB does not buy remaining mechanical life

How you value that 636,000 THB is where the decision actually turns. In Option 3, the mechanism, bearings, ball screws, linear guides, servo motors and control panel components are all new. In Option 2, the mechanism that has been running since 2012 stays exactly where it is, with a new jig and a new unit bolted on top.

So choosing Option 2 means saving 636,000 THB in exchange for accepting whatever service life remains in a mechanism installed in 2012. If positioning accuracy problems appear right after the modification, or the main spindle assembly needs renewal a few years later, that saving disappears immediately. Conversely, where the mechanism is in good condition and there is a defensible view of how many years it has left, Option 2 is a rational choice.

What that judgement requires is measurement rather than intuition. Backlash. Repeat positioning accuracy. Wear on the main sliding surfaces. Servo load factors. Measure those while you are collecting modification quotes and the question of whether 636,000 THB is worth paying can be argued internally on evidence. The procedure for evaluating replacement and modification side by side, machine by machine, is set out in legacy equipment replacement.

The Option 1 versus Option 2 gap can be closed starting now

There is a second thing worth taking from this model. The 880,000 THB gap between Option 1 and Option 2 is not a difference in machine performance. Same machine, same modification requirement. The only variable is whether the documentation survived.

And of that 880,000 THB, 600,000 THB sits in the survey and software layers – the cost of finding out. If updated drawings, commented source code, a parameter backup and the change history are written into the handover conditions of this modification, that cost does not arise next time. What you specify as deliverables while you are still evaluating quotes is what sets the price of your next modification.

Issues that apply specifically in Thailand and ASEAN

Relocated machines have a forked drawing set

A substantial share of the equipment in Japanese-owned plants in Thailand was moved here from a plant in Japan. Relocation involves disassembly for shipping and reassembly on site, adaptation to local supply voltage, and substitution of locally procured components. The changes made during that work are frequently never reflected in the original drawings held in Japan.

The result is a drawing set that has forked into two versions. There is the original held by the design department in Japan, and there is the redlined, fitted-to-reality version held on the shop floor in Thailand. Issue only the first when requesting quotes and a discrepancy will surface after work begins. Issue the second, or both. Confirming which you have is an internal task that can be finished before the enquiry goes out.

Local parts availability drives lead time and therefore downtime

Modifications regularly call for components from the same manufacturer and the same series as what is already installed – another axis from the same servo amplifier family, another sensor of the same model as the existing ones. The question that then matters is whether that model number is held in distributor stock inside Thailand.

If it is, it arrives in days. If it is not, it is imported from Japan or China, and customs clearance time is added on top of the lead time. That timeline feeds straight into fixing the downtime window. Entering a stoppage without all parts on site extends the stoppage by exactly as long as the wait. Recall how heavily downtime weighed on the all-in figures above. Procurement routes need to be locked down at the same time as the specification, not after it.

Match the modification to the maintenance capability you actually have

The people maintaining the machine after the modification are the local maintenance team, not engineers flying in from Japan. The specification should therefore be set within what that team can support. Concretely, that means keeping added devices with manufacturers already present on the machine, providing Thai language on the HMI screens, and producing the operating manual and work standards in Thai.

Neglect this and the modification succeeds while uptime falls. When a fault occurs, nobody can recover the machine and the line waits on an enquiry to Japan. The training and Thai-language work sits inside the safety and documentation layer cost.

Design the stoppage schedule around the public holiday calendar

Thailand’s public holidays differ from Japan’s, and periods such as the Songkran break offer consolidated windows in which a line can be down anyway. When the modification downtime goes into the production plan, whether it can be overlaid on one of those periods changes the lost production figure materially. Even a modification needing six days generates almost no lost profit if it lands on an extended holiday. Schedule it into a peak period and the model’s lost production applies in full. Because lost production drives the modify-or-replace decision, schedule design deserves the same weight in your planning as price negotiation.

Ten things to settle on paper before requesting a quote

Below are the items to write down internally before the modification enquiry goes out. Fill these in and quote accuracy improves while post-award variations fall.

Item 10 deserves a note. Obtaining a new-build quote purely to get the modification approved looks like extra work and is in fact a shortcut. Once you know whether the modification lands at 42.4% or at 69.1% of the new build, the internal debate finishes on numbers. Without a new-build quote, nobody in the room can say whether the modification price is high or low.

Five common failure patterns

Requesting quotes without first checking what records exist

The most frequent failure by a wide margin. Send the enquiry to several companies without checking whether drawings and program exist, and you receive a low quote written on the assumption that they do alongside a high quote written on the assumption that they do not. The plant selects the low one, discovers after award that the records are missing, and absorbs both a variation and a longer stoppage. Open the cabinet and inspect the contents before the enquiry goes out.

Leaving downtime out of the production plan

The modification cost reaches the capital request but the downtime never reaches the production schedule. Start negotiating the window only after approval and the project sits with parts delivered and no date to install them. Secure the window in parallel with collecting quotes, not afterwards.

Going to other vendors without asking the original machine builder

If the original builder can take the work, the survey layer is minimised, because the design intent already exists inside that company. Even where a Japanese builder cannot support the job directly, it is often still possible to obtain the source code and drawings from them. Skipping that enquiry means paying for reading effort that was never necessary.

Dropping the safety rebuild out of the quote

The reasoning goes that adding a jig does not change anything about safety, so the safety and documentation layer comes out of the scope. But adding a jig changes how an operator’s hands enter the machine, and adding a screw-tightening unit introduces a new hazard. The risk assessment gets repeated, the operating manual and work standards get revised, and training happens. That is the layer budgeted at 180,000 THB to 220,000 THB in the model.

Accepting the machine back without collecting the updated records

The machine runs, acceptance is signed, and neither updated drawings nor source code changes hands. At the next modification, the identical survey cost is paid a second time. In the terms of this model, a job that should have qualified as Option 1 reverts to Option 2 conditions. The difference is 880,000 THB. Put handover deliverables into the acceptance criteria.

Frequently asked questions

How much does an equipment modification cost

It depends on content, but for the model case in this article – one semi-automatic assembly machine gaining one jig set and one screw-tightening unit – the figure is 1,400,000 THB where drawings and program are available and 2,280,000 THB where they are missing. What sets the range is not the volume of hardware being added but whether the documentation for the existing machine survived. When requesting quotes, ask for the survey, mechanical, electrical and control, software, safety and documentation, and installation and commissioning layers separately so the numbers can be compared.

Which is cheaper, production line modification or a new build

On quoted cost alone, modification. In the model case, a modification with complete records is 42.4% of the new build and one without records is 69.1%. The gap narrows once lost production is included. The modification without records comes to 2,856,000 THB all-in against 3,492,000 THB for the new build, a difference of 636,000 THB. Since that difference does not buy remaining mechanical life, a new build can be the rational answer depending on the condition of the machine.

What do we do about retrofitting existing machinery when there are no drawings

Modification is still possible without drawings. It does, however, require redrawing the schematics, analysing the program and extracting the parameters, and those hours land on the survey layer. In the model case, the survey layer rises from 120,000 THB to 520,000 THB. One workable approach is to define the scope of investigation before anything else and order the survey on its own as a first phase. Taking the full modification quote after the survey results are in keeps post-award variations down.

Can we modify only the control system

Yes. Leaving the mechanism untouched and reworking only the PLC program and HMI screens is a common form of control system modification. Requirements such as automating changeover, adding variant selection or recording production counts can often be met inside the software layer alone. The precondition is that the source code and the device allocation table for the existing program still exist. Without source, the entire existing logic has to be read in order to guarantee that current behaviour is preserved, which takes more hours than new development.

Does modifying equipment make it a legally new machine

This article does not identify any legislation applying directly to a factory in Thailand or Japan that provides that modification creates a new machine. Regulation (EU) 2023/1230 carries the concept that a party making a substantial modification may be regarded as the manufacturer, but that is an EU regulation and does not apply directly to a plant in Thailand or Japan. It bites when machinery or product is shipped to Europe, and when a European parent company standard is applied internally. Risk assessment under Article 28-2 of Japan’s Industrial Safety and Health Act includes the point at which equipment is changed among its triggers, but it is a duty of effort, with certain areas such as chemical substances having been made mandatory.

How much downtime should we plan for

It depends on how much can be settled in advance. The model case uses 3 days for the modification with drawings and program available and 6 days without them. With records in hand, logic is built in the office and the stoppage is limited to fitting, wiring and functional checks. Without them, the investigation itself happens inside the stoppage. Fix the lost profit per day of downtime first and the difference in days can be compared in money.

How do we explain internally whether a modification quote is expensive

Obtain the new-build quote in parallel and rank both on quoted cost plus lost production. Once you can state the modification as a percentage of the new build, the discussion closes on numbers. Then present the quote split between the survey layers and the build layers, and where the survey layer is large, explain it as the cost of having lost the information about that machine. Making that argument leads naturally into managing handover deliverables from then on.

What should we collect after the modification is finished

Updated mechanical drawings and electrical schematics, commented PLC source code, HMI screen data, a servo parameter backup together with the reasoning behind the settings, and the change history for this project. Put those five into the acceptance conditions and the survey layer cost does not arise at the next modification. In the model case, the survey and software layers alone account for a 600,000 THB difference.

Summary

Equipment modification quotes diverge because of a difference in surviving information about the machine, not a difference in the volume of work. For a machine whose records are gone, a substantial share of the modification price is the cost of re-establishing what is there rather than the cost of building anything. In the model case, the identical requirement produced 1,400,000 THB and 2,280,000 THB, and of the 880,000 THB gap, 600,000 THB was the increase in survey and software analysis.

Convert downtime into money and rank the options on the all-in figure, and the gap between the modification without records at 2,856,000 THB and the new build at 3,492,000 THB narrows to 636,000 THB. That difference does not buy the remaining life of a mechanism that has been running since 2012. So the decision becomes explainable in numbers by running the four gates of information, structure, safety and downtime in order, and by reading cost across the five layers of survey, mechanical, electrical and control, software, and safety and documentation.

One final point. Write this project’s deliverables into the acceptance conditions and the next modification starts from Option 1 conditions. Cost control for equipment modification begins the moment the current modification ends.

If the evidence needed to choose between modification and new build is not yet assembled internally, or if you would rather start by establishing how much documentation for the existing machine still survives, you are welcome to raise it through our contact page. We support both on-site machine surveys within Thailand and the preparation of layer-by-layer estimates.

References

• Regulation (EU) 2023/1230 (Machinery Regulation) – EU-OSHA
• Regulation (EU) 2023/1230, full text (EUR-Lex)
• Ministry of Health, Labour and Welfare, “Revision of the Comprehensive Safety Standards Guidelines for Machinery” (Notification No. 0731001, 31 July 2007)
• Ministry of Health, Labour and Welfare, Occupational Safety and Health Information Site, “Risk Assessment”
• Ministry of Health, Labour and Welfare, “Notification of Machinery Hazard Information Has Become a Duty of Effort” (Ordinance on Industrial Safety and Health, Article 24-13, as amended)
• Thailand’s industrial output slips 3.1 pct in June (Xinhua, 27 July 2026)
• Thai June factory output falls 3.1% y/y, weaker than forecast (Business Recorder)