The invoice from the third-party warehouse arrives every month, and the amount creeps up year after year. The racks on site are full, and pallets are starting to line the aisles. At some point you decide something has to change, you start searching for an automated warehouse price, and the results are useless: one page says a few million yen, the next says hundreds of millions of yen, and both appear on the same screen. This article works through what actually sets the order of magnitude, which lines of a quotation deserve your attention, and how the payback moves in a model calculation built around a Japanese-owned plant in Thailand, in the order you would tackle it in practice.
Why warehouse automation is on the agenda now
Automated storage is not new technology. The reason it has suddenly become a board-level topic at Japanese-owned plants in Thailand and the wider ASEAN region comes from two directions at once: the market and the cost of labour.
The Southeast Asian warehouse automation market is projected to reach USD 1.63 billion by 2031
According to Mordor Intelligence, the Southeast Asian warehouse automation market is projected to grow from USD 0.81 billion in 2025 to USD 0.91 billion in 2026, and to USD 1.63 billion by 2031. The CAGR for 2026 to 2031 is 12.36%.
Before you copy that anywhere, pin down what the figure covers. This is the whole of Southeast Asia and the whole of warehouse automation. It is not Thailand alone, and it is not automated storage and retrieval systems (AS/RS) alone. Lose that scope when you transfer it into an internal deck and it quietly becomes a different claim, such as “the Thai automated warehouse market is worth USD 1.63 billion”. That is exactly the kind of line a capital request gets challenged on first.
Inside the total, the mobile robots segment held a 29.76% share in 2025, worth USD 0.24 billion, and is expected to grow at a CAGR of 13.92% through 2031, ahead of the market average. In other words, even as the whole market expands, the faster growth is in equipment you can move, not in large fixed installations.
By country, Indonesia holds the largest share at 28.63%, while Vietnam is the fastest growing at a CAGR of 13% through 2031. For Thailand, the report points to the concentration of the automotive industry and the tax incentives of the EEC (Eastern Economic Corridor) as tailwinds. It also notes that AS/RS adoption is concentrated in automotive and pharmaceutical cold chain, that is, in applications involving heavy loads or temperature control.
Read the other way round, this means AS/RS is still not the default answer for an ordinary parts warehouse that handles neither heavy loads nor controlled temperatures. That observation feeds directly into the lighter alternatives discussed later in this article.
Thailand’s minimum wage holds at THB 400 per day
Now the labour side of the equation. In Thailand, a revision effective 1 July 2025 set the minimum wage for all industries within Bangkok at THB 400 per day, up from THB 372. Hotels rated two stars or above, or with 50 rooms or more, and their related facilities are covered in every province. In the manufacturing belt, four provinces and one district, including Chonburi and Rayong, have applied THB 400 since January 2025.
No further revision took effect in 2026, and the range of THB 337 to THB 400 per day remains in force. On top of that, a phased increase in the ceiling used to calculate social security contributions began in January 2026.
So the headline daily rate is flat, while the total cost carried by the employer is drifting upwards. In operations such as warehousing, where output scales with headcount, that gap compounds.
Vietnam raised wages by roughly 7.2% on average in January 2026
For companies running production across several ASEAN sites, the Vietnamese picture matters too. Under Decree 128/2025/ND-CP, regional minimum wages rose by approximately 7.2% on average with effect from 1 January 2026.
| Region | Monthly (VND) | Hourly (VND) | Increase |
|---|---|---|---|
| Region 1 | 5,310,000 | 25,500 | +7.06% |
| Region 2 | 4,730,000 | 22,700 | +7.26% |
| Region 3 | 4,140,000 | 19,900 | +7.25% |
| Region 4 | 3,700,000 | 17,800 | +7.25% |
The increase ranges from 7.06% to 7.26% depending on the region. It is not “7.2% everywhere”; 7.2% is the average. That looks like a trivial distinction until you build a side-by-side comparison of several countries, at which point rounding differences turn into reconciliation problems.
Even so, “labour is expensive” will not carry the capital request
Having spent three sections on wages, here is the counterpoint, and it is the central message of this article.
The system type determines the order of magnitude of an automated warehouse price; the payback is driven not by headcount savings but by storage density (that is, cancelling external warehouse space) and by BOI incentives.
The model calculation later in this article shows it in numbers: if headcount reduction is the only benefit on the sheet, the payback period comfortably exceeds ten years for every system type. Rising labour cost is a legitimate trigger for looking at automation, but it is not the engine of the return. Hold that distinction from the start and you will read quotations differently.
The automated warehouse price is set by system type
The single most important fact to absorb before you request a quotation is this: the system type determines the order of magnitude of an automated warehouse price, long before the detailed specification comes into play.
Price bands run from a few million yen to hundreds of millions
Itoh Denki’s technical column sets out the price bands by system type as follows.
| System type | Indicative price band (Japan domestic) | Typical use |
|---|---|---|
| Vertical lift module | From a few million yen | Smallest-scale deployments |
| Mini-load (bucket/tote type) | From tens of millions of yen | Small cases, high-mix small parts |
| Large shuttle and unit-load (pallet) systems | From hundreds of millions of yen | High-density storage of heavy or large products |
Between a vertical lift module and a large pallet system there is close to a thousandfold spread. A gap that wide cannot be closed by trimming the specification to fit the budget. The only workable sequence is to fix the system type first, then refine the specification.
APT’s guidance adds a figure for the entry level: JPY 10,000,000 to JPY 20,000,000 for a small system, made up of four elements, namely storage equipment, stacker crane, control equipment and inventory management system. The same source states plainly that the figure “varies widely depending on the installation environment, scale and system configuration”. When you research price levels, that caveat is usually the most honest sentence on the page.
The system type follows the load format
Type selection is not driven by preference or by budget. It is driven by what you are storing. The main types behave as follows.
- Pallet AS/RS: for heavy and large products. Its strength is high-density storage that uses the full building height.
- Bucket AS/RS: for small cases and irregular items. Suited to raising picking accuracy across a high-mix range of small parts.
- Free-size AS/RS: for warehouses handling cartons, containers and other goods of varying shape and size. Typically considered for high-mix low-volume operations.
- Vertical lift and shuttle systems: configurations that exploit height where floor area is limited. These sit well with phased deployment.
If, picturing your own inventory, the answer is “we have pallets, we have totes, and occasionally long items”, forcing that mix through a single machine will send the price sharply upward. Classify the load formats first and draw a line between what you automate and what stays manual. Done in that order, the result is usually cheaper.
Put the advantages and the drawbacks on the same page
The same column lists the advantages of automated storage as better storage density, better work efficiency, lower labour cost and higher inventory accuracy, and the drawbacks as the required upfront investment, the risk exposure when equipment fails and reduced flexibility.
A common failure in capital requests is a document that lists only the advantages. Decision makers will ask about the downside without fail, and an inability to answer on the spot reads as shallow preparation. Equipment failure risk and loss of flexibility are precisely the two points the shop floor cares about most. How many hours to recover when the system stops? Will the rack pitch still fit when the product mix changes? Put both questions to the vendors at the enquiry stage.
Do not carry yen price levels straight into Thailand
Every yen figure above is an indication for the Japanese domestic market. Procure in Thailand and freight, local installation, local building works for power and fire protection, and customs conditions all change.
Use the yen figures as an indication of the order of magnitude only. “Vertical lift modules sit in the millions of yen, mini-load systems in the tens of millions, large pallet systems in the hundreds of millions” is a fair takeaway. “It costs JPY 10,000,000 in Japan, so it costs the same in Thailand” is not. That is also why the model calculation later in this article is built in Thai baht.
Break the quotation into five cost lines
Once the type is settled, the quotation follows. The outcome to avoid above all others is a quotation that reads “automated storage system, one lot” on a single line. One line cannot be benchmarked against a competitor, gives you nothing to negotiate against, and leaves you with no basis for argument when a variation order appears later.
At the enquiry stage, tell every vendor that you want the price split into the following five cost lines. Whether or not you say that one sentence changes the quality of everything that follows.
Cost line 1: core equipment (stacker crane, racking, conveyors)
The heart of the system: stacker cranes or shuttles, racking, and the infeed and outfeed conveyors. In the model calculation below, this line accounts for roughly 72.2% of the investment.
The two things to nail down here are the number of storage locations and the throughput. Until those are fixed, no vendor can confirm machine count or model. Put differently, until you have decided them yourself, no amount of chasing will produce an accurate quotation.
Cost line 2: installation, steelwork and anchoring
The work of setting the equipment on site and fixing it down. The figure moves depending on whether you are retrofitting an existing building or building new. In an existing building you need to confirm floor load capacity, where anchors can be placed, and the delivery route, including whether large components physically fit through the building openings.
This line cannot be priced accurately without a site survey. Installation costs quoted from drawings alone frequently rise once someone measures the building. When you compare quotations, make sure every vendor has been on site. Comparing a pre-survey estimate with a post-survey quotation and concluding that vendor A is cheaper is not a comparison at all.
Cost line 3: control panels, host software and WMS integration
The control panels that run the machinery, the warehouse management system (WMS) that sits above them, and the interface to your existing ERP or production management system.
This line has the widest spread between vendors and the greatest tendency to grow after the fact. The reason is simple: at enquiry stage, nobody can see what “integrate with the existing system” actually contains. Which data, in which direction, at which moment, at what level of granularity? Leave that vague, accept an “integration, one lot” price, and variations appear the day detailed design begins.
Seen as a question of plant-wide infrastructure, this overlaps heavily with the issues worth settling before you push ahead with factory automation: the design object is the flow of information, not the individual machine. Treat an automated warehouse as a system that holds inventory data rather than as a large shelf, and you will read the quotation correctly.
Cost line 4: building modifications, power and fire protection
Automated storage uses height, so the building itself gets touched. Insufficient clear height means roof work. Taller racking means additional sprinklers and a review of fire protection. More machinery means a review of incoming power capacity.
This line is often absent from the equipment vendor’s quotation. A single sentence reading “building works excluded” is not unusual. A quotation that looked cheap but simply excluded the building works is the most common accident in any comparison. When a quotation lands, check for this line first.
Cost line 5: spares, training and commissioning support
Stock of spare parts, training for operators and maintenance staff, and support during start-up. It is the smallest line by value, and cutting it is felt later.
Spares matter especially in Thailand. If a control board or a motor has to be shipped from Japan, the lead time from failure to recovery becomes warehouse downtime, hour for hour. Which parts, how many, held locally is a decision about availability, not a bargaining chip in price negotiation.
Costs that never appear in the quotation but always occur
Here is the part that matters. The five lines above appear on the vendor’s quotation. There are also costs you will incur on your own side that never appear on any quotation. If they are not in the capital request, someone will say later that they were never told.
First, modifying your existing WMS or ERP. The inventory side of your existing landscape has to be adapted to the automated warehouse. That work goes to the incumbent system vendor, so it is not in the equipment vendor’s price. If the existing system is old, or the change request comes back expensive, this becomes the bottleneck for the whole plan.
Second, the effort to redesign locations. Which items go into the automated warehouse, which stay on manual racking, and how the location numbering scheme is structured. This cannot be outsourced wholesale. It falls to the people who know the items, squeezed between their normal duties. Treat that effort as free and the project will slip. Formally allocate a share of someone’s time.
Third, training time. Not only operating instructions, but abnormality handling, daily checks and basic recovery procedures, delivered to both shifts if you run two. The production time lost while people are being trained is also a cost.
Fourth, parallel running at start-up. Immediately after cutover you will run the old method alongside the new one for a period. During those weeks headcount goes up, not down. The model calculation assumes benefits from year one; in reality it is prudent to assume that the first few months deliver zero benefit, or a negative one.
Seven variables that move the automated warehouse price
Two quotations for the same “mini-load system” can differ enormously depending on conditions. Give every vendor the same answers to the following seven points and their quotations become directly comparable.
1. Load format and weight
Pallet, bucket or case, or irregular? Mixed? What is the maximum weight per unit load? This determines the system type itself.
2. Number of storage locations
How many pallet positions, how many case locations? Without this, no price can be produced, ever. Undecided location counts are the single biggest reason quotations come back as one lot.
3. Throughput
Give the peak movements per hour, not the average. Design to the average and you will jam at month end or on a specific weekday. Overstate the peak and crane count rises, taking the price with it. This has to come from measurement.
4. Clear height and building constraints
Retrofit or new build? What is the usable clear height? Column pitch, floor load capacity, delivery route. The tighter the retrofit constraints, the further you move from standard machine specifications and the higher the price.
5. Temperature range
Ambient, controlled ambient, chilled or frozen? The lower the temperature, the more the insulation, condensation control and equipment specification change, and the higher the price. AS/RS adoption is well advanced in pharmaceutical cold chain partly because keeping people working in temperature-controlled space is expensive.
6. Operating hours and redundancy
24-hour operation calls for standby units or duplicated equipment. “We can wait until the morning shift” and “one hour of downtime stops the line” imply completely different levels of redundancy. Leave this vague, pick the cheapest quotation, and you will regret it after go-live.
7. Interfaces to host systems
What connects to the existing ERP, WMS and production management systems, and how? Which side owns the master data? Which side holds the authoritative stock figures? The integration method, whether file transfer, API or direct database access, is decided at the same time as the machine selection, not after it.
Consolidate those seven points into a single requirements sheet before you go out to enquiry and the quality of what comes back changes. Send an enquiry with three or more of them blank and every vendor will price defensively. That is not bad faith on their part; with no information, it is the only rational response.
Payback model | comparing two logistics automation scenarios
Now to the numbers. Everything below is an original model calculation prepared for this article, not the actual results of a specific project. Change the assumptions and the conclusion changes, so substitute your own figures and recalculate.
Shared assumptions (model case)
- Subject: a Japanese-owned parts plant in Chonburi Province, Thailand, with a finished goods and parts warehouse.
- Current state: fixed racking, forklifts and manual picking, with 12 warehouse operators across two shifts.
- Labour cost: fully loaded employer cost per warehouse operator is set at THB 18,000 per month. This is built from the Bangkok minimum wage of THB 400 per day multiplied by 26 days per month, giving THB 10,400 per month, then multiplied by roughly 1.73 to cover overtime, social security, benefits and administrative overhead. It is an assumption; actual figures differ by company, so replace it with your own.
- External warehouse: inventory that does not fit on site is stored externally at THB 180,000 per month (THB 2,160,000 per year).
- Corporate income tax rate: calculated at 20%, the standard Thai rate.
- Headcount treatment: the reduction is not redundancy. It is the hiring that is avoided through natural attrition and redeployment, and only that is booked as a benefit.
On these assumptions, current warehouse labour cost is 12 people times THB 18,000 times 12 months, or THB 2,592,000 per year in the model. Add the THB 2,160,000 of external warehouse charges and the warehouse function is absorbing THB 4,752,000 a year.
Scenario A: mini-load AS/RS installed in one step
| Cost line | Amount (THB) | Share |
|---|---|---|
| Core equipment (stacker crane, racking, conveyors) | 26,000,000 | approx. 72.2% |
| Installation, steelwork and anchoring | 4,000,000 | approx. 11.1% |
| Control panels, host software, WMS integration | 3,000,000 | approx. 8.3% |
| Building modifications, power, fire protection | 2,000,000 | approx. 5.6% |
| Spares, training, commissioning support | 1,000,000 | approx. 2.8% |
| Total | 36,000,000 | 100% |
Everything other than the core equipment adds up to THB 10,000,000, roughly 27.8% of the investment. Budget on the basis of “what does the machine cost” alone and you will be nearly a third short.
Annual benefits are as follows.
| Benefit | Amount (THB/year) | Share | Basis |
|---|---|---|---|
| Labour reduction (12 to 7 people, 5 positions) | 1,080,000 | approx. 29.7% | 5 x 18,000 x 12 |
| Full cancellation of external warehouse (storage density x2.0) | 2,160,000 | approx. 59.3% | 180,000 x 12 |
| Fewer mis-shipments, less stocktaking effort | 400,000 | approx. 11.0% | assumption |
| Total | 3,640,000 | 100% | – |
Against that, running costs increase after go-live.
| Additional running cost | Amount (THB/year) | Basis |
|---|---|---|
| Maintenance contract (3% of investment) | 1,080,000 | 36,000,000 x 0.03 |
| Additional electricity | 180,000 | assumption |
| Total | 1,260,000 | – |
Netting the two, the annual net benefit is 3,640,000 minus 1,260,000, or THB 2,380,000.
- Simple payback: 36,000,000 / 2,380,000 = approx. 15.1 years
- Cumulative net benefit over 3 years: THB 7,140,000 (19.8% of the investment)
- Cumulative net benefit over 10 years: THB 23,800,000, giving a 10-year ROI of approx. -33.9%
What happens with BOI incentives? Here we use a simplified approach: the exemption cap, at 50% of the investment or THB 18,000,000, is deducted from the investment, and the calculation runs on an effective net investment of THB 18,000,000 (the conditions under which this simplification holds are set out below).
- Simple payback: approx. 7.6 years
- 10-year ROI: approx. +32.2%
One caveat is essential. Using the full THB 18,000,000 cap requires THB 90,000,000 of taxable income over three years (18,000,000 / 20%), or THB 30,000,000 a year on average. Fall short of that level and the benefit shrinks proportionately.
Scenario B: vertical lift modules plus AGVs, deployed in phases
| Cost line | Amount (THB) | Share |
|---|---|---|
| Vertical lift modules, 2 units | 6,000,000 | approx. 42.9% |
| AGVs, 3 units (including charging equipment) | 4,500,000 | approx. 32.1% |
| WMS modification, handheld terminals, location setup | 2,500,000 | approx. 17.9% |
| Floor works, power, safety fencing | 1,000,000 | approx. 7.1% |
| Total | 14,000,000 | 100% |
Annual benefits are as follows.
| Benefit | Amount (THB/year) | Share | Basis |
|---|---|---|---|
| Labour reduction (12 to 9 people, 3 positions) | 648,000 | approx. 32.8% | 3 x 18,000 x 12 |
| Cancellation of half the external warehouse (storage density x1.4) | 1,080,000 | approx. 54.6% | 90,000 x 12 |
| Fewer mis-shipments, less stocktaking effort | 250,000 | approx. 12.6% | assumption |
| Total | 1,978,000 | 100% | – |
| Additional running cost | Amount (THB/year) | Basis |
|---|---|---|
| Maintenance contract (3% of investment) | 420,000 | 14,000,000 x 0.03 |
| Additional electricity | 80,000 | assumption |
| Total | 500,000 | – |
The annual net benefit is 1,978,000 minus 500,000, or THB 1,478,000.
- Simple payback: 14,000,000 / 1,478,000 = approx. 9.5 years
- Cumulative net benefit over 3 years: THB 4,434,000 (31.7% of the investment)
- Cumulative net benefit over 10 years: THB 14,780,000, giving a 10-year ROI of approx. +5.6%
- With BOI incentives (exemption cap THB 7,000,000): effective net investment of THB 7,000,000, payback approx. 4.7 years, 10-year ROI approx. +111.1%
Here too, using the full cap requires THB 35,000,000 of taxable income over three years, roughly THB 11,670,000 a year.
What the comparison shows
| Item | Scenario A (one step) | Scenario B (phased) | Difference |
|---|---|---|---|
| Investment (THB) | 36,000,000 | 14,000,000 | 22,000,000 |
| Annual benefit (THB) | 3,640,000 | 1,978,000 | 1,662,000 |
| Additional annual running cost (THB) | 1,260,000 | 500,000 | 760,000 |
| Annual net benefit (THB) | 2,380,000 | 1,478,000 | 902,000 |
| Simple payback | approx. 15.1 years | approx. 9.5 years | approx. 5.6 years |
| Cumulative net benefit, 3 years (THB) | 7,140,000 | 4,434,000 | 2,706,000 |
| 3-year cumulative / investment | 19.8% | 31.7% | 11.9 points |
| 10-year ROI | approx. -33.9% | approx. +5.6% | 39.5 points |
| Payback with BOI (simplified) | approx. 7.6 years | approx. 4.7 years | approx. 2.9 years |
| 10-year ROI with BOI (simplified) | approx. +32.2% | approx. +111.1% | 78.9 points |
Note: the difference column is the difference between the rounded figures shown above.
Three conclusions follow from this table, in order.
First, a difference of THB 22,000,000 in investment buys a difference of only THB 902,000 in annual net benefit. Recovering the extra THB 22,000,000 out of the extra THB 902,000 would take approximately 24.4 years in the model. Judged on its own, the incremental portion of the single-step deployment is a poor use of capital.
Second, in both scenarios the largest benefit is not labour but the cancellation of external warehouse space. In A it is 2,160,000 out of 3,640,000, or 59.3% of the benefit; in B it is 1,080,000 out of 1,978,000, or 54.6%. Labour reduction accounts for only about 29.7% in A and about 32.8% in B.
That is precisely why building the case on headcount savings alone produces numbers detached from reality. Divide Scenario A’s THB 36,000,000 investment by the THB 1,080,000 of labour savings and the payback works out at more than 33 years. For Scenario B, THB 14,000,000 divided by THB 648,000 is approximately 21.6 years in the model. Never divide an automated warehouse price by labour savings alone.
Third, whether or not BOI incentives are included moves Scenario B’s 10-year ROI from +5.6% to +111.1%. The incentive scheme is not a bonus you hope to pick up; it changes the denominator of the investment decision itself. The correct sequence is to confirm BOI eligibility and then decide on the investment. Reverse it and you will be redoing the decision.
What you must understand about BOI before using these figures
The calculations above use a simplified approach: the exemption cap is deducted from the investment to arrive at an effective net investment. That treatment carries a condition.
The BOI measure for improving production efficiency is not a grant covering 50% of the investment. What is exempted is corporate income tax, and the cap on the total exemption is 50% of the investment. A company with low taxable income therefore cannot use the full cap. In a loss-making year there is no corporate income tax to exempt at all, so the benefit is zero.
Scenario A needs THB 90,000,000 of taxable income over three years to use its THB 18,000,000 cap; Scenario B needs THB 35,000,000 to use its THB 7,000,000 cap. Whether you can meet that condition is a question about your own profit and loss, not about the equipment. Confirm it with finance and tax before, not after, the investment decision.
Use BOI incentives to move the denominator
BOI’s automation-related incentives come in two forms: a general measure open to all industries, and announcements targeted at specific sectors. The conditions and the scope are entirely different, so do not conflate them. Internal documents that mix the two are common.
General measure: production efficiency improvement (automation)
This is the general category relevant to most manufacturers.
| Item | Content |
|---|---|
| Corporate income tax exemption | 3 years |
| Cap on exempted amount | 50% of the investment |
| Condition for raising the cap | Cap rises to 100% if 30% or more of the automation machinery and equipment is manufactured in Thailand |
| Minimum investment (general companies) | THB 1 million (excluding land and working capital) |
| Minimum investment (SMEs) | THB 500,000 (excluding land and working capital) |
| SME criteria | Annual revenue of THB 500 million or less, and Thai shareholding of 51% or more |
Examples of eligible equipment include automatic weighing machines, automatic packaging machines, automatic sewing machines, automated guided vehicles (AGVs), sensor systems and equipment for making plant assets IoT-capable.
Two points deserve attention. One is that the minimum investment of THB 1 million for general companies and THB 500,000 for SMEs is not a high bar. This is not a scheme reserved for major capital projects such as a full AS/RS. The other is that AGVs are explicitly named among the examples of eligible equipment, which means a phased deployment along the lines of Scenario B can also be considered for incentives.
The condition that the cap rises to 100% when 30% or more of the automation machinery and equipment is manufactured in Thailand feeds straight into the procurement plan. Import everything from Japan, or combine locally manufactured equipment? That choice can double the exemption cap, which makes it worth considering at the design stage.
Announcement 4/2569: additional incentives for the automotive industry
Announcement 4/2569, published in the Royal Gazette on 31 March 2026, provides additional incentives restricted to the automotive industry.
| Item | Content |
|---|---|
| Royal Gazette publication | 31 March 2026 |
| Scope | General automotive manufacturing (category 3.6) and PHEV/HEV manufacturing (category 3.8) |
| Incentive | 50% corporate income tax exemption for 3 years on investment in automation and robotics systems (excluding land and working capital) |
| Condition for raising the cap | Rises to 100% where 30% or more of the machinery value contributes to Thailand’s domestic automation industry |
| Minimum investment | THB 1 million (excluding land and working capital) |
| Application deadline | End of 2027 |
| Other | Includes exemption from import duty on machinery |
The critical point is that this announcement applies to operators falling under categories 3.6 and 3.8, and not to all industries. Making automotive components does not automatically mean your business category qualifies. Whether it does has to be checked against the terms of your own BOI promotion. Write “we are automotive, so this incentive applies” into a capital request and you will be issuing a correction later.
For those who do qualify, the application deadline of end-2027 starts to bite. An investment like an automated warehouse, which takes time from study to purchase order, leaves less slack than it first appears when you work backwards from that date.
Get the sequence right
The most common accident with incentive schemes is sequence.
Approval must be obtained before the machinery is procured and imported for the import duty exemption to apply. Starting the application after the order is placed, or after the equipment reaches the port, is too late. The incentive is lost in full.
In practice, once you have narrowed the system type and have a budget-level figure, start confirming BOI eligibility and the required documentation. Do it in parallel with machine selection, not immediately before the order. The 90-day roadmap later in this article places this confirmation in the final phase for exactly that reason.
Material handling equipment to consider before an AS/RS | start with in-plant logistics improvement
Everything so far has been about automated storage, but in practice there are several options worth checking before you consider an AS/RS at all. As noted, the market research finds AS/RS adoption concentrated in automotive and pharmaceutical cold chain. For an ordinary parts warehouse, lighter measures frequently deliver a better return on capital.
Automating transport with AGVs and AMRs
If the warehouse problem is “not enough people to move things” rather than “not enough space to store things”, transport automation is the first thing to look at. AGVs (automated guided vehicles) and AMRs (autonomous mobile robots) can be introduced without touching the building, and capacity can be adjusted by adding units, which makes them a natural fit for phased deployment.
As cited earlier, within the Southeast Asian warehouse automation market the mobile robots segment held a 29.76% share in 2025 (USD 0.24 billion) and is expected to grow at a CAGR of 13.92% through 2031, ahead of the 12.36% market average. The money is flowing towards equipment you can move rather than towards large fixed installations.
For the decision criteria and day-to-day operational issues, see our discussion of what to weigh up when considering AGVs and AMRs at a plant in Thailand. Since AGVs are named among the examples of eligible equipment under the general BOI measure, the incentive angle can be considered alongside.
Mobile racking
Mobile racking raises storage density by eliminating fixed aisles. The racks themselves travel on rails, opening an aisle only where and when it is needed. The investment is an order of magnitude below an AS/RS and it fits readily into an existing building.
It is not suited to fast-moving items, because of the wait while the aisle opens. Where slow-moving inventory is consuming the space, it is a high-return option.
Vertical lift modules
The standard answer where floor area is tight and clear height is available. The price band starts from a few million yen, the lightest category of automated storage. Typical uses are dense storage of small parts, tooling and some die and mould inventory.
Scenario B selects vertical lift modules for exactly this “start small” character. Install one unit, confirm that the operation works, then add units or move on to a different system type.
Location redesign and in-plant logistics flow improvement
And then there is the measure that often delivers the most for zero capital investment.
Has anyone run an ABC analysis on the current warehouse? Are the highest-frequency items sitting on the racks furthest from the door? Is dead stock occupying prime locations, year after year? Do forklift routes cross each other and create queuing?
Rearranging locations, disposing of dead stock and redrawing the traffic routes routinely improves both storage density and work efficiency on their own. More importantly, this work is required in any case if you do install an automated warehouse. Migrate to an AS/RS without first cleaning up the location scheme and you simply automate the confusion.
Looking at how the flow itself should be designed, the same logic runs through digital transformation in Southeast Asian logistics: considering the upstream and downstream processes together, rather than carving out the warehouse and automating it in isolation, often ends up requiring less capital.
A safe sequence is: location redesign, then light equipment such as mobile racking or vertical lift modules, then AGVs and AMRs, and only then an automated warehouse. Start from the AS/RS and you risk applying an order-of-magnitude larger solution to a problem that could have been solved far more cheaply.
Four common ways this goes wrong
Four classic missteps in automated warehouse projects. All of them are expensive to reverse.
Mistake 1: going to enquiry without a storage density target
The most common of all. Send “we are considering an automated warehouse, please provide a budget figure” and no price can be produced while the number of storage locations is undefined. Vendors will make assumptions and return a budget figure, but each vendor’s assumptions will be different.
The result is a set of one-lot quotations that cannot be compared. You cannot tell whether a price gap reflects specification, assumptions or margin. The only way back is to level the conditions and re-issue. Deciding the location count and throughput up front is faster in the end.
Mistake 2: writing the capital request on labour savings alone
As the model calculation showed, using headcount reduction as the sole benefit produces payback periods detached from reality: more than 33 years for Scenario A and approximately 21.6 years even for Scenario B, in the model.
Connect the benefits to line items where cash actually moves: external warehouse charges, the cost of handling mis-shipments (return freight, expedited replacement), and line stoppage time for stocktaking. These are documented in invoices and records, so finance can follow them. “Work becomes easier” and “less time spent searching” do not translate into money.
Mistake 3: leaving WMS and ERP integration until later
Warehouses where the equipment was installed but inventory data ended up maintained in two places, with people reconciling it by hand, genuinely exist. Which system is authoritative, the automated warehouse or the existing inventory system? When is it updated? How are discrepancies resolved? Go live without deciding, and the labour savings you assumed will not materialise.
Decide the integration specification at the same time as the machine selection. Take the “choose the machine first, think about systems later” route and you will be bending the systems around the machine’s constraints.
Mistake 4: starting the BOI application after placing the order
As noted, approval must be obtained before the machinery is procured and imported for the import duty exemption to apply. Get the sequence wrong and the incentive disappears.
What makes this one nasty is how late the discovery comes. The order is placed, the equipment is delivered, and only when finance goes to file the application does anyone learn that it no longer qualifies. By then the denominator behind the investment decision has already changed. In Scenario B, an investment expected to return +111.1% over ten years becomes one that returns +5.6%.
A 90-day evaluation roadmap
Finally, a way to structure the evaluation across 90 days. Equipment studies have a habit of consuming a year while information is gathered, so working to a deadline is the more reliable approach.
Day 0-15: measure the current state
Do not call vendors yet. Put your own current state into numbers first. Five things to measure.
- Storage locations: how many pallet positions and case locations are in use today, including those at the external warehouse
- External warehouse cost: the monthly charge, together with the contract term and cancellation conditions
- Peak throughput: movements per hour at peak, not the average
- Mis-shipments: the number of cases, and the cost of handling each one (freight, expedited supply)
- Stocktaking effort: annual man-hours for stocktaking and the associated line stoppage time
These 15 days are the most valuable work in the whole 90. Without measurement here, every subsequent decision rests on guesswork.
Day 16-30: set targets and write a one-page requirements sheet
Using the measured values, set target values for storage density and throughput, in the form of “1.3 times the current number of storage locations” or “XX movements per hour at peak” (set the figures from your own measured data).
Then consolidate the seven variables described earlier (load format, location count, throughput, clear height and building constraints, temperature range, operating hours and redundancy, host system interfaces) onto a single A4 page. A thick document is not required. If it does not fit on one page, it means something has not been decided yet.
Day 31-50: shortlist two system types and go to enquiry on identical terms
Working from the requirements sheet, narrow the field to two system types. One leaves nothing to compare; three or more scatters the evaluation.
Issue the enquiry by giving every vendor the same single page, stating location count, throughput, temperature range and the systems to be interfaced. Level the conditions here and the quotations come back comparable. At the same time, ask for the price split into the five cost lines described earlier: core equipment; installation and steelwork; control, host software and WMS integration; building, power and fire protection; spares, training and commissioning.
Day 51-70: re-sort the quotations by cost line and compare
Rearrange the quotations you have received line by line. Check that every vendor has included the same lines, and in particular that building, power and fire protection are included. A quotation that excludes them will always look cheaper than one that includes them.
At the same time, put running costs and maintenance terms side by side: annual maintenance fee, coverage hours, response time to attend site, spare parts supply arrangements, and the guaranteed parts supply period. The model calculation assumed maintenance at 3% of the investment, but real terms differ by vendor. This is the point at which you look at differences in terms, not only in price.
Your own internal costs (existing WMS modification, location redesign effort, training, parallel running at start-up) are also built up at this stage.
Day 71-90: confirm BOI and taxable income, then decide
In the final phase, confirm BOI eligibility and the taxable income outlook with your tax function.
There are three things to check. First, whether the planned equipment can qualify for the incentive. Second, whether taxable income over three years is likely to be sufficient to absorb the full exemption cap. Third, whether the schedule from application to approval fits the timing of machinery procurement and import.
Only when those three are in place is the denominator of the investment decision fixed. Concluding “payback is ten years, so we pass” before the denominator is settled is a premature decision. In Scenario B of the model, the presence or absence of BOI moved the payback from approximately 9.5 years to approximately 4.7 years.
Frequently asked questions
How much does an automated warehouse cost?
The system type determines the order of magnitude. As an indication for the Japanese domestic market, vertical lift modules start from a few million yen, mini-load (bucket type) systems from tens of millions of yen, and large shuttle and unit-load (pallet) systems from hundreds of millions of yen. A figure of JPY 10,000,000 to JPY 20,000,000 is also cited for a small system, comprising storage equipment, a stacker crane, control equipment and an inventory management system.
These are Japanese domestic indications, however, and procurement in Thailand involves different conditions for freight, installation and local building works. Use them as an indication of the order of magnitude and confirm the real figure by going to enquiry on your own conditions. The model calculation in this article assumes THB 36,000,000 for a single-step mini-load deployment in Thailand and THB 14,000,000 for a phased deployment of vertical lift modules plus AGVs, both of which are assumptions.
What is the difference between an automated warehouse and a WMS?
An automated warehouse is physical equipment that stores and moves goods; a WMS (warehouse management system) is software that manages inventory information. They perform different roles, and neither one completes the job alone.
What matters in practice is that, on top of these two, you also need interfaces to your existing ERP and production management systems. The model calculation allocates THB 3,000,000 to “control panels, host software, WMS integration” in Scenario A and THB 2,500,000 to “WMS modification, handheld terminals, location setup” in Scenario B. In Scenario B that is roughly 17.9% of the investment. Budget on the machine price alone and this part goes missing.
Can a small or medium-sized plant do this?
There are options scaled to the site. Vertical lift modules can be considered from a price band starting at a few million yen, and combining them with AGVs allows a phased approach. Scenario B in this article is modelled on exactly that premise.
On the incentive side, the BOI production efficiency improvement measure sets a minimum investment of THB 1 million for general companies and THB 500,000 for SMEs, both excluding land and working capital. The SME criteria are annual revenue of THB 500 million or less and Thai shareholding of 51% or more. This is not a scheme available only to large capital projects.
That said, the smaller the operation, the more the taxable income constraint bites. What is exempted is corporate income tax, so a company with limited taxable income cannot absorb the full cap. For that reason a phased approach that keeps the investment itself small tends to work out better.
Do BOI incentives apply to automated warehouses?
Examples of eligible equipment under the production efficiency improvement measure (automation) include automatic weighing machines, automatic packaging machines, automatic sewing machines, automated guided vehicles (AGVs), sensor systems and equipment for making plant assets IoT-capable. Whether a specific installation qualifies depends on its configuration, so confirm with BOI and with your own accounting and tax advisers before finalising the investment plan.
Two points must be understood in the process.
First, this is not a grant. What is exempted is corporate income tax, and the cap on the total exemption is 50% of the investment (or 100% where 30% or more of the automation machinery and equipment is manufactured in Thailand). A company with low taxable income cannot use the full cap.
Second, sequence. Approval must be obtained before the machinery is procured and imported for the import duty exemption to apply. Start the application after placing the order and the incentive disappears.
Separately, for the automotive industry, Announcement 4/2569, published in the Royal Gazette on 31 March 2026, offers a 50% corporate income tax exemption for 3 years on investment in automation and robotics systems, with an application deadline of end-2027, for general automotive manufacturing (category 3.6) and PHEV/HEV manufacturing (category 3.8). This announcement is limited to the automotive categories and is separate from the general measure open to all industries.
How long does implementation take?
Manufacturing and installation lead times vary widely with system type and scale, so this is a question for vendors at the enquiry stage. What this article can offer is an indication of the time from study to investment decision.
Specifically, a 90-day approach: measure the current state in Day 0-15; set targets and produce a one-page requirements sheet in Day 16-30; shortlist two system types and go to enquiry on identical terms in Day 31-50; compare quotations line by line in Day 51-70; and confirm BOI eligibility and taxable income before making the investment decision in Day 71-90.
Note that the BOI application schedule is often the real constraint. Because procuring and importing machinery before approval breaks the condition for the import duty exemption, the application-to-approval period has to be worked backwards from the intended order date.
Summary
The key points of this article on automated warehouse price and payback.
- The system type determines the order of magnitude of the price. As a Japanese domestic indication: vertical lift modules from a few million yen, mini-load (bucket type) systems from tens of millions of yen, large shuttle and unit-load (pallet) systems from hundreds of millions of yen. A small system is also cited at JPY 10,000,000 to JPY 20,000,000. Conditions in Thailand differ, so treat these as an indication of the order of magnitude.
- Ask for the quotation split into five cost lines: core equipment; installation and steelwork; control, host software and WMS integration; building, power and fire protection; spares, training and commissioning. In Scenario A of the model, everything other than the core equipment accounted for roughly 27.8% of the investment. On top of that, existing WMS modification, location redesign effort, training and parallel running at start-up never appear on a vendor’s quotation yet always occur.
- Labour is not the engine of the payback. In the model, the largest benefit in both scenarios was the cancellation of external warehouse space (59.3% of the benefit in A, 54.6% in B). Divide by labour savings alone and Scenario A takes more than 33 years. Never divide an automated warehouse price by labour savings alone.
- More investment does not deliver proportionally more net benefit. A difference of THB 22,000,000 in investment produced a difference of THB 902,000 in annual net benefit. Simple payback was approximately 15.1 years (A) versus approximately 9.5 years (B), and 10-year ROI approximately -33.9% (A) versus approximately +5.6% (B), all in the model.
- BOI incentives move the denominator itself. With BOI, Scenario B’s 10-year ROI moves from +5.6% to +111.1%. But what is exempted is corporate income tax; 50% of the investment is not paid out as a grant. Using the full cap requires substantial taxable income over three years (THB 90,000,000 for A, THB 35,000,000 for B). And without approval before the machinery is procured and imported, the incentive is lost.
- Look at the lighter measures before the automated warehouse. A safe order is location redesign and flow improvement, then mobile racking or vertical lift modules, then AGVs and AMRs, then an automated warehouse. In the Southeast Asian warehouse automation market too, the mobile robots segment is expected to grow at a CAGR of 13.92% through 2031, ahead of the 12.36% market average.
When the external warehouse invoice starts to feel heavy, the first thing to do is not to collect quotations. It is to establish your current number of storage locations and your target number. Fix those two and a price can be produced; leave them open and every vendor you approach will return a one-lot figure.
Talk to us
If you are working through warehouse automation or an automated storage project, get in touch through the TOMAS TECH contact form. There is no expectation that you engage us to install the equipment. “We want to agree the location count and targets before we go to enquiry”, “we want to judge whether we really need an automated warehouse or whether something lighter will do”, “we want to work out whether the general BOI measure or the automotive announcement applies to us” are all perfectly good reasons to call. Working from Bangkok, we have supported automation and system implementation for Japanese-owned manufacturers, and we are happy to help you structure the questions around your own conditions.
References
- Southeast Asia Warehouse Automation Market (Mordor Intelligence)
- What is an automated warehouse? Types, advantages, disadvantages and implementation cost (Itoh Denki, in Japanese)
- The purpose and cost of implementing an automated warehouse system (APT, in Japanese)
- BOI investment promotion policy for automation machinery and robotics industries in Thailand (BOI, Japanese-language document)
- Thailand Unveils New Incentives for Automotive and HEV/PHEV Manufacturing (Tilleke & Gibbins)
- Bangkok minimum wage raised to THB 400 per day (JETRO, in Japanese)
- Minimum wage raised by an average of approximately 7.2% in January 2026 (Vietnam / JILPT, in Japanese)
