Intended readers: Executives, site managers, plant managers, production engineers, and administrative staff of Japanese manufacturers based in Thailand
When thinking about Thailand’s manufacturing sector in 2026, the first thing to grasp is that this is not a simple story of “the economy heading uniformly downward.” The reality is somewhat more complex. Overall growth in the Thai economy is expected to slow, and there is uncertainty around export orders and the tourism recovery. At the same time, areas attracting investment—digital, electrical and electronics, EVs, green manufacturing, data centers, and renewable energy—enjoy a clear tailwind.
In other words, the decision facing Thailand’s manufacturing sector in 2026 is not “stop investing or move forward.” More accurately, it is about separating “investments that should be stopped” from “investments that should be advanced now.”
For Japanese manufacturers in particular, what matters is reframing capital and system investment not as “surplus investment made in good times” but as “defensive investment to protect profit margins.” Labor costs, electricity costs, logistics costs, the cost of quality response, a shortage of managers, over-reliance on individuals, and inventory volatility—these remain on the floor whether times are good or bad. And when growth slows, this waste squeezes profits all the more strongly.
In this article, taking into account the situation in Thailand in 2026, we organize which investments Japanese manufacturers should stop and which they should advance, from the perspectives of IoT, automation, AI, accounting DX, and BOI incentives.
1. Thailand’s manufacturing sector in 2026 is “weak, but not finished”
The World Bank’s Thailand Economic Monitor projects that Thailand’s GDP growth rate will slow to 1.6% in 2026. Behind this lie weak global trade, high household debt, and a sluggish tourism recovery. While the Thai economy continues to grow, this is no longer a phase where the whole is lifted with the momentum of the past.
Turning to manufacturing, while there are moments when S&P Global’s Thailand Manufacturing PMI holds in expansion territory, weakness in new export orders has been noted. This does not mean that Thailand’s domestic production floors will come to a complete halt. However, it shows that the premises of “if you make it, it sells,” “add people and it runs,” and “even with some waste, sales will absorb it” are breaking down.
What is troublesome for Japanese companies is that in such an environment, investment approval from headquarters tends to become cautious. When the economic outlook is weak, the decision tends to be “let’s hold back on investment for now” or “let’s wait and see until next term.” From the floor’s perspective, however, labor shortages, increasingly sophisticated quality requirements, inventory accuracy, equipment downtime, and a shortage of managers will not wait.
What is needed here is to separate economic judgment from investment judgment. Stopping large-scale expansion investment because the economy is weak is a reasonable decision. On the other hand, stopping even improvement investment because the economy is weak is dangerous. The reason is that if you stop investment that protects profit margins, then in a phase where sales do not grow, only costs and fixed expenses grow heavier.
2. Investments to stop: large-scale system investment with vague objectives
Let us first consider the investments that should be stopped. A representative example is large-scale system investment with vague objectives.
For example, there are cases where a company tries to introduce ERP, MES, WMS, BI, and an IoT foundation all at once, solely for reasons such as “headquarters says DX,” “other companies are starting smart factories,” or “we want to renew our old systems.” Of course, these mechanisms themselves are important. However, if they are introduced without a clear picture of the floor’s improvement themes or a path to recouping the investment, the system becomes an “unused box.”
At Thai sites, there is also the circumstance that business understanding, language, and expectations easily diverge among Japanese headquarters, local management, Thai staff, and outside vendors. The Japanese side demands “standardization,” while the local side asks to “fit it to the current floor.” The vendor hears the requirements and builds accordingly, but does not go so far as to ensure operational adoption. When a large system is introduced within this structure, the floor often cannot master it after introduction, and Excel and paper continue to remain.
Therefore, what should be stopped in 2026 is not DX itself. It is DX where “which floor numbers to improve” has not been decided.
For example, the following kinds of investment should be paused once.
- IoT investment where the purpose of introduction is only “visualization” and no improvement actions have been decided
- ERP renewal that systematizes today’s complex operations as-is without organizing existing business processes
- Daily-report and quality-record systems that do not consider the burden of on-floor data entry
- AI introduction where the premise for recouping the investment is the vague notion that “effects will appear if people use it”
- BI builds whose sole purpose is creating reports for Japanese headquarters
These are investments that tend to fail even in good times. All the more in a phase like 2026, where slowing growth and rising costs are felt simultaneously, vague investments become difficult to justify internally.
3. Investments to advance: small-scale automation that protects profit margins
On the other hand, there are investments that should be advanced now. These are small-scale automation efforts that protect profit margins.
The automation referred to here is not necessarily a large-scale line lined with robots. Rather, what is realistic for Thailand’s mid-sized Japanese manufacturers are the following areas.
- Automating data entry that people transcribe every day
- Digitizing paper daily reports, quality records, and inspection sheets
- Automatically capturing equipment operating hours and reasons for stoppage
- Making inventory quantities, work-in-process, and receipt/dispatch records real-time
- Making inspection results and complaint information searchable
- Automatically summarizing meeting and report content and turning it into tasks
Each of these looks unremarkable on its own. However, they reliably reduce on-floor management costs, errors, rework, and waiting time. What matters in a phase of weak growth is increasing the profit that remains from existing sales, without depending on sales expansion. In that sense, small-scale automation is an offensive investment and, at the same time, a defensive investment.
In Thailand especially, rising labor costs, recruiting difficulties, and a shortage of managers are ongoing challenges. The model of solving problems by adding people has become harder than before. It is not that the floor cannot run because there are not enough people; rather, you need to redesign operations so that they run even with fewer people.
What is important here is not to “automate everything.” What should be automated first are tasks that occur every day, involve no judgment, are prone to error, and affect downstream processes. For example, work where paper production daily reports are transcribed into Excel, that Excel is tallied at month-end, and it is further processed for headquarters reporting. Such work is a heavy burden for both the floor and the administrative department, and moreover is too slow to use for management decisions.
The criteria for the investment decision are simple. How many hours per month will the automation cut? How much will errors and rework be reduced? Will it lead to early detection of equipment stoppages and defects? Can it be recouped within three years? Investment that can answer these questions is worth advancing even in a phase of weak growth.
4. Don’t let IoT end at “visualization”
A common failure in manufacturing DX is stopping at “visualization.”
Displaying equipment utilization, downtime, production counts, defect counts, and power consumption on a dashboard is an important first step. However, the floor does not change merely because numbers appear on a screen. Rather, if no one is designated to look at them, they are not used in meetings, and improvement owners cannot translate them into action, the dashboard becomes mere decoration.
If you are going to advance IoT investment, the first thing to decide is not the type of sensor. What you should decide is “who does what after looking at the numbers.”
For example, if you are capturing equipment downtime, you need to classify the reasons for stoppage, review the top factors weekly, and have maintenance, production engineering, and floor leaders decide on improvement actions. If you are capturing defect rates, you must link them to lots, processes, operators, and equipment conditions, and record them at a granularity usable for root-cause analysis. If you are visualizing inventory, it is meaningless unless it connects to reorder points, work-in-process, stocktaking discrepancies, and shipping schedules.
In other words, IoT is not “investment to capture data” but “investment to change the conversation about improvement.”
At Japanese plants in Thailand, problem perception sometimes diverges between Japanese managers and Thai floor leaders. The Japanese manager feels “why isn’t it improving,” while the floor side feels “I don’t know what criteria to prioritize by.” When there is shared data here, the conversation changes. That is because they can talk based on downtime, defect counts, workload, and inventory discrepancies rather than someone’s gut feeling.
In that sense, the condition for IoT investment success lies in operational design rather than technology. In which meeting do you look at it? Which KPI do you improve? Who is the owner? How do you record improvement results? Only when you design this far does visualization lead to improvement.
5. BOI incentives are not “something to look into later” but something to build into investment design
When considering investment in automation, IoT, AI, data analytics, and enterprise management systems in Thailand, BOI incentives are an important point of discussion. The BOI Investment Promotion Guide 2025 sets out a framework in which automation and robotics; AI, machine learning, and big-data analytics; software and IT systems for enterprise management; cloud services; and the like are treated as investment and expenditure.
Of course, which investments actually qualify varies depending on the nature of the business, the investment plan, application conditions, and existing BOI status. Therefore, the final judgment requires confirmation with specialists or the BOI. But what is important as a management decision is not to treat BOI as “something to look into after the investment is decided to see whether it can be used.”
BOI should be built in from the early stages of investment design.
For example, the persuasiveness of internal justification differs depending on whether you submit a proposal as a mere equipment update or organize it as an investment plan that includes smart manufacturing, efficiency improvement, and data utilization. If you design it as a “productivity-improvement investment” that includes IoT sensors, equipment-integration software, the cloud, AI analytics, ERP integration, and even floor training, you can explain it not as a mere cost but as investment aligned with policy direction.
BOI’s recent direction leans toward higher value-added, smart manufacturing, sustainability, and the use of digital technology, rather than low-value-added simple processing. The Thai government too recognizes that competitiveness is hard to maintain with the conventional labor-intensive model alone. This is an important message for Japanese manufacturers as well.
In other words, considering BOI incentives is not just about looking at tax benefits. It is also about confirming whether your own investment aligns with the Thai government’s industrial policy. Investment that aligns with policy direction is easier to justify in internal proposals and also leads to long-term competitiveness.
6. Plant DX without accounting DX does not lead to management decisions
When people hear “plant DX,” their attention inevitably turns to equipment, sensors, and floor apps. However, what ultimately matters to executives is how floor data connects to profit.
Utilization went up. Defect rates went down. Inventory accuracy improved. These are important. But unless they connect to product-level cost, customer-level gross margin, project-level profitability, and faster monthly closing, they remain difficult to use for management decisions.
At Japanese manufacturers in Thailand, it is not uncommon for floor data and accounting data to be disconnected. The floor holds production results in Excel, accounting makes journal entries in a separate system, inventory is managed separately by warehouse staff, and yet another set of reporting materials is created for Japanese headquarters. In this state, executives cannot immediately judge “which product is truly profitable” or “which process is eroding profit.”
In a phase like 2026, where growth slows and rising costs are felt, it is dangerous to look only at sales. What is needed is to look at gross margin and cash. In which product does profit remain, with which customer does effort pile up, and in which process do losses occur? To make this visible, you need to consider accounting DX and floor DX together, not separately.
For example, connect equipment-operation data, defect data, work hours, material usage, and inventory discrepancies with cost accounting on the accounting side. Make it possible to view billing, purchasing, inventory, and production results end-to-end. Grasp profitability—which previously was known only at month-end—on a weekly, and ideally daily, basis. With such a mechanism, executives can act sooner.
Plant DX is not only for making the floor’s work easier. It is for enabling executives to judge “where to invest next and where to stop.” For that, a perspective that connects floor data and accounting data is indispensable.
7. Rather than replacing people, start AI where it supplements a shortage of managers
Interest in using generative AI and AI agents is rising rapidly in manufacturing as well. However, when using AI in a plant, you do not need to start suddenly with advanced demand forecasting or autonomous control. Rather, where effects appear most easily first is in areas that supplement a shortage of managers.
At Thai sites, challenges easily arise such as a Japanese manager overseeing multiple departments, incomplete delegation of authority to Thai managers, inconsistent granularity in reporting, and post-meeting actions falling through the cracks. These are, before any advanced AI, problems of information organization and follow-up.
AI demonstrates its power here.
- Summarizing meeting audio and extracting decisions and tasks
- Picking up anomalies and unhandled items from daily reports
- Searching past cases of quality complaints
- Organizing equipment-trouble histories and presenting similar causes
- Organizing reports in Japanese, Thai, and English across multiple languages
- Semi-automatically creating weekly reports for headquarters
This is not AI for reducing headcount. It is AI for letting managers concentrate on the issues they should originally be addressing. At Japanese companies in particular, reporting/contact/consultation, meeting minutes, approvals, and follow-up greatly affect work quality. If AI can assist here, it increases managers’ time and stabilizes communication with the floor as well.
What is important is not to let AI introduction end as a one-off experiment. What do you have the AI read? Who checks the summaries it produces? Where are tasks registered? How is completion confirmed? Only when you design this far does AI take root in operations.
8. Frame investment decisions in terms of “three-year payback”
To get DX investment approved at headquarters or in management meetings, explaining the technology alone is not enough. What is needed is an explanation of the payback on the investment.
When explaining to Japanese headquarters from a Thai site in particular, “it becomes convenient,” “we can visualize,” and “we can use AI” are weak. What headquarters wants to know is how much is invested, which costs go down, which risks are reduced, and when it can be recouped.
One useful benchmark here is three-year payback. Of course, the appropriate period varies by industry and equipment content. However, investment whose effects can be explained within three years tends to carry persuasive power even in a phase of weak growth.
For example, 100 hours of transcription work per month can be cut. Stocktaking discrepancies decrease. Equipment downtime decreases by several hours a month. Scrap from defects decreases. Monthly closing speeds up. Complaint-handling time is shortened. You convert these into monetary terms and see whether the investment amount can be recouped in three years.
Furthermore, there are effects that are hard to monetize. Resolving over-reliance on individuals, delegating authority to local talent, responding to quality audits, restoring customer trust, and improving the accuracy of headquarters reporting. These are hard to see as direct reduction amounts, but they are extremely important for business continuity. Therefore, in the investment proposal it is good to explain “quantitative effects” and “risk-reduction effects” separately.
9. Concrete investment themes to advance in 2026
Taking the discussion so far into account, the investment themes that Japanese manufacturers in Thailand should prioritize in 2026 can be organized as follows.
1. Combined introduction of operational visualization and improvement meetings
Build a mechanism that captures equipment operation, reasons for stoppage, production counts, and defect counts, and uses them in weekly improvement meetings. What matters is not just building a dashboard but clarifying the improvement actions and the owners.
2. Digitizing paper forms and daily reports
Digitize paper daily reports, inspection sheets, and quality records so they can be searched, tallied, and approved. A design that does not increase the floor’s data-entry burden is important. Combining smartphones, tablets, QR codes, and voice input makes adoption easier.
3. Visualizing inventory, work-in-process, and receipts/dispatches
Inventory accuracy is directly linked to manufacturers’ profit margins. To reduce excess inventory, stockouts, stocktaking discrepancies, and stagnant work-in-process, you need to connect floor movement with ERP. Even just starting from barcodes or QR codes has an effect.
4. Unifying quality data and complaint response
For quality issues, the speed of response after they occur determines trust. By linking lots, processes, inspection results, work records, and customer complaints, root-cause investigation and recurrence prevention become faster.
5. Accounting DX and cost management
To connect the results of floor improvement to management decisions, cost and gross margin need to be visible. Make profitability visible by product, by customer, and by process, and build a structure that judges by profit rather than sales.
6. Meeting-minutes AI and reporting/communication AI
A mechanism that uses AI to organize meetings, daily reports, and reports and turn them into tasks is effective against a shortage of managers. At Japanese sites where Japanese, Thai, and English are mixed, multilingual support also becomes a major value.
7. An automation and AI investment plan premised on BOI
Rather than thinking about automation, AI, data analytics, enterprise management IT, and cloud utilization separately, organize them as an investment plan aligned with BOI’s policy direction. This makes them easier to justify on both the internal-proposal and external-system fronts.
10. The direction of support TOMAS TECH envisions
The value TOMAS TECH should provide to Japanese manufacturers in Thailand is not merely building systems. It is turning floor challenges into mechanisms that connect to management decisions.
Like contract development, merely hearing out requests and building exactly that can end up simply transferring the floor’s reliance on individuals into the system. What is needed from here is support premised on standardization, non-customization, operational adoption, and data utilization.
For example, start with one line, one process, one form, and one meeting. Capture data there, use it in improvement meetings together with the floor, and confirm the effect. Once the effect is visible, roll it out horizontally to other processes and other sites. With this approach, you can reliably take root on the floor while keeping investment risk down.
Also, what is important for Japanese companies is bridging Japanese headquarters and the Thai floor. Headquarters judges by numbers and risk. The floor judges by day-to-day usability and burden. To satisfy both, you need a design that understands not only technology but also business, accounting, floor training, language, and culture.
What TOMAS TECH should aim for is not just flashy phrases like “AI works on its own around the clock,” but actually picking up the floor’s reports, organizing the issues, connecting them to the next action, and creating a state in which executives can make decisions. It should be AI, IoT, and accounting DX for making the manufacturing floor win.
Summary: The weaker the economy, the more the quality of investment is tested
Thailand’s manufacturing sector in 2026 is not an environment to advance through with optimism alone. Growth is slowing, there is weakness in export orders, and labor, logistics, and quality-response costs grow heavier. But that is not a reason to stop all investment.
What should be stopped is large-scale investment with vague objectives. What should be advanced is investment that reduces floor waste, protects profit margins, supplements a shortage of managers, and speeds up management decisions.
Don’t let IoT end at visualization; connect it to improvement actions. Start automation not only with large-scale lines but with paper, transcription, inventory, and quality records. Rather than replacing people, use AI starting where it reclaims managers’ time. Accounting DX is needed to convert floor improvement into profit. And BOI incentives should be built in from the early stages of the investment plan.
The weaker the economy, the more a company’s differences appear not in the amount invested but in the quality of investment. For Japanese manufacturers in Thailand, 2026 is not merely a year of endurance. It is a year to strengthen the floor, protect profit margins, and build the mechanisms to prepare for the next growth phase.
For its part, TOMAS TECH should provide practical support that connects IoT, automation, AI, accounting DX, and BOI, so that Japanese manufacturers in Thailand can advance not “vague DX” but “floor-driven DX that pays back in three years.”
