Walk into the shipping area of almost any factory in Thailand and you will see the same picture. One operator erects the boxes, another packs the product, a third tapes the lid, a fourth applies the label, and two more lift cartons onto a pallet. The instinctive reaction is to ask a vendor for a quote on the machine that replaces the hardest job. But packaging line automation succeeds or fails on something else entirely: whether the chain from case erecting to the outbound conveyor is designed as one system. This article looks at packaging line automation from a systems integration angle, covering how to find the real bottleneck, how layout choices differ, how to run a request for proposal, and how to think about payback.
What packaging line automation actually means
The packaging process is a chain of seven stations
What we call a packaging line is normally a chain of the following steps.
- Case erecting, where a flat corrugated blank is formed into a box and the bottom is sealed
- Case packing, where product is collated and dropped into the box at a fixed count
- Sealing, where the top flaps are closed with tape or hot melt adhesive
- Labelling, where product codes, barcodes, destinations and lot data are applied
- Bundling or shrink wrapping, where several boxes are grouped or covered with film
- Palletizing, where boxes are stacked onto a pallet in a defined pattern
- Outbound conveying, where pallets are moved to a stretch wrapper or a shipping bay
The point is that these are not seven independent jobs. The exit of each station is the entrance of the next. If the case erector can only produce ten boxes a minute, no case packing robot downstream will push the line past ten boxes a minute. Packaging line automation is the act of designing that flow, not the act of making individual machines faster.
Optimising single machines does not speed up the whole
The most common decision on the floor is to replace only the station that consumes the most labour. Palletizing is usually first in line, because the loads are heavy and the ergonomic cost is obvious. Choosing a palletizing robot is often the right call, and how to specify and price one is covered in our article on palletizing robot implementation.
What happens next, though, is predictable. Sealing and labelling upstream stay manual, the palletizer waits for boxes, and the equipment sits idle for a large share of the shift. Headcount barely drops, lead time barely moves, and the investment has already been spent.
The difference is not machine performance. It is whether the stations are connected. How the line behaves when one station stops, where buffers sit, and who switches which machine in what order during a format change are the questions that separate an integrated line from a collection of separate machines.
Where the market is heading
The packaging robot market is moving in the same direction. Technavio forecasts the market to grow at a compound annual growth rate of 10.9% between 2025 and 2030, with the market opportunity increasing by USD 5.59 billion. The regional split matters more than the headline. Asia Pacific is expected to drive 41.2% of that growth, which puts Thailand and the wider ASEAN region at the centre of the shift.
Demand is moving fast as well. According to the Association for Advancing Automation, robot orders from the North American food and consumer goods sector jumped 105% year over year in the third quarter of 2025 alone. Thailand and the wider ASEAN region are not tracking that exact figure, but the underlying driver, a tight labour market pushing manual packaging lines toward automation, is the same one Thai plants are facing.
How to find the real bottleneck
The three numbers to measure first
Integrated design starts with data, not intuition. Deciding that “case packing is slow” by feel is how investment ends up in the wrong place. Measure at least these three figures per station.
- Effective throughput, meaning boxes actually produced per hour, measured on the floor rather than taken from a catalogue
- Downtime and its breakdown across format changeover, material replenishment, jams, quality checks and waiting
- Assigned headcount and the share of their time spent on real work versus waiting
Record these for even one week and the bottleneck becomes obvious. In most plants the true constraint is not the station with the most people, but the station that stops most often for changeover or replenishment.
Typical bottlenecks by station
Each station fails for different reasons.
| Station | Typical bottleneck | How to spot it |
|---|---|---|
| Case erecting | Frequent blank replenishment stops the machine | Replenishment dominates the downtime breakdown |
| Case packing | Product collation is unstable and product backs up | Boxes or product accumulate before the station |
| Sealing | Tape run-out and poor adhesion require rework | A person is permanently assigned to fix seals |
| Labelling | Label stock and data are swapped by hand for each variant | Changeover time is far longer than at other stations |
| Bundling and shrink wrapping | Heating time is fixed and cannot be shortened | Product always accumulates upstream while downstream runs empty |
| Palletizing | Changing the stacking pattern takes a long time | Output drops only on format change days |
| Outbound conveying | Pallets wait for a free shipping bay | The line stops at the same time every afternoon |
If two or more rows describe your plant, a single machine purchase will not fix it. The connections and buffers between stations are where the work has to happen.
The format count problem
One constraint has grown quickly in recent years: the number of formats a line has to handle. Industry analysis notes that a system designed for four formats a decade ago may now be running twelve or more. More stock keeping units means more box sizes, more counts per box, and more stacking patterns.
Manual changeover takes anywhere from 20 minutes to over an hour depending on complexity. A line that changes format three times a day is losing between one and three hours of production to changeover alone. Very few plants include this figure when hunting for bottlenecks, and it is often the single largest source of loss.
Automation islands versus an integrated line
Four hidden losses
When automated machines sit scattered without being connected to each other, the result is what integrators call automation islands. Four losses come with that state, and none of them appear in a quotation.
First, people are needed between the islands. If the case erector sits three metres from the case packer, somebody has to move boxes across the gap or at least watch it.
Second, the effect of a stoppage becomes unpredictable. An integrated line slows the downstream section or absorbs the interruption in a buffer. A set of islands keeps running blind until an operator notices.
Third, data fragments. Counting output station by station gives no single view of effective line throughput or of where the minutes were lost, which leaves improvement work without evidence.
Fourth, changeover becomes personal knowledge. Each island switches differently, the order matters, and eventually only one person can do it.
When an integrated line is worth it
Integration is not always the answer. The table below sets out the decision.
| Factor | Phased single machines fit | Integrated line design fits |
|---|---|---|
| Volume | One shift with spare capacity | Two or more shifts with routine overtime |
| Format count | Few and stable | Many variants changing often |
| Space | Existing layout cannot be altered | The shipping area can be rearranged |
| Budget | Must fit a single fiscal year | A multi-year capital plan exists |
| Staffing | Headcount is currently secured | Recruitment is difficult and unlikely to improve |
| Quality | Visual inspection is sufficient | Traceability records are required |
When the choice is genuinely unclear, design the whole line first and invest in stages. If installation positions, conveyor heights and control interfaces are already fixed, later additions do not trigger rework. Designing the whole and investing in parts is the most practical route in nearly every case.
The semi-automatic option
Industry analysis of packaging trends points out that fully automated systems deliver the largest long-term labour savings but demand heavy upfront capital, complex implementation and long downtime when they fail. Semi-automated systems cost less, install faster, and combine with existing equipment with minimal disruption. They assist operators rather than replacing them, with the equipment carrying the load while the operator controls positioning.
For many plants in Thailand, the sensible sequence is to stabilise the flow semi-automatically, collect throughput and downtime data, and only then decide which stations justify full automation.
Layout design for a packaging line
Straight, L-shaped or U-shaped
Layout follows the shape of the shipping area and the outbound path.
A straight layout gives the clearest flow and the shortest transport distance, and is the first choice wherever a long clear run is available along a wall. Its drawback is that infeed and outfeed sit at opposite ends.
An L-shaped layout uses a building corner and works around existing columns, at the cost of turntables or curved conveyor at the bend, which adds jam risk.
A U-shaped layout brings infeed and outfeed close together, shortening operator walking distance and making it easier for one person to supervise several stations. It usually needs more floor area than a straight run.
Buffers and accumulation
What really decides whether an integrated line works is conveyor design. With enough accumulation between stations, the upstream can keep running while the downstream is stopped for tens of seconds. With no buffer at all, one stoppage propagates instantly through the whole line.
As a rule of thumb, provide accumulation between stations equal to the average recovery time of the station that stops most often. If labelling recovers in an average of 40 seconds, place a conveyor long enough to hold 40 seconds of boxes ahead of it. The mechanics of that design are covered in our article on conveyor design and material handling.
Build changeover into the specification
Because format count is becoming the dominant constraint, write these three requirements into the specification at layout stage.
- Tool-free adjustment, so guide widths and stop positions move with levers or handwheels
- Recipe management, so every machine switches together when a product code is selected on the panel
- A fixed switching sequence displayed on screen, so the order does not depend on one person
The mechanism selection for the case packing station itself, including robot end effector types, is covered in our article on case packing automation robots and is not repeated here.
Choosing an integrator and running the request for proposal
Packaging line integration is where integrator capability shows most
Buying a single machine only requires comparing vendor specifications. Integrating a packaging line means binding machines from several manufacturers into one control architecture, designing the transport between them, and defining behaviour on failure. Integrator experience shows up directly in the result. General selection criteria are set out in our article on how to choose a robot system integrator; the points below are specific to packaging lines.
What to state in the request for proposal
| Item | What to state in the request for proposal |
|---|---|
| Products and all formats | Box dimensions, weight, count per box and annual volume share for every format |
| Required capacity | Target throughput and how many consecutive hours it must be held |
| Operating conditions | Shift pattern, operating days and planned stoppage windows |
| Changeover | Number of format changes per day and the acceptable changeover time |
| Site constraints | Column positions, ceiling height, floor loading, power and compressed air capacity |
| Upstream and downstream | How product arrives and the path to the shipping bay |
| Data requirements | Output records, downtime records, traceability fields and where they are sent |
| Acceptance criteria | Factory acceptance test pass conditions, continuous run duration and yield |
The last row matters most. If acceptance only requires that the equipment runs, a line that cannot hold its rated capacity in real production will still pass. Require every format to be run and a target throughput to be sustained for a stated number of consecutive hours.
What to look for when comparing quotations
When comparing quotations, check that control integration and host system interface hours are priced, that conveyor and buffer lengths match the stated capacity, that commissioning duration and manpower are explicit, that a spare parts list and local service response time are included, and that operator and maintenance training is offered in a language the site actually uses.
Cost structure and payback
Read the cost in four layers
Looking only at machine list prices guarantees the budget will overrun. Build the number in four layers instead.
| Cost layer | What it contains | Commonly missed |
|---|---|---|
| Equipment | Station machines, robots and peripheral units | Spare grippers and change parts are usually quoted separately |
| Transport and integration | Conveyor, buffers, guarding, control panels, integration software | Integration software hours vary most between bidders |
| Installation | Mechanical installation, wiring, air piping, floor work, relocation | Removal of existing equipment is frequently omitted |
| Commissioning | Trial running, training, early yield loss, initial spare stock | Lost production during ramp-up is rarely costed |
A single machine quotation covers only the first layer. On an integrated line the transport and integration layer is a substantial share of the total, and it is precisely that layer which delivers the labour saving and uptime a set of islands cannot.
What belongs in the payback calculation
Automating the end-of-line steps, case packing and palletizing, can pay back in 12 to 24 months in high-volume, high-overtime environments. Two conditions deserve attention here: “end-of-line” and “high-volume, high-overtime.” Extending automation upstream to labeling and shrink-wrapping changes both the investment and the payback, so this figure should not be applied to the whole line without adjustment; build the payback from each station instead. On a single shift with spare capacity, the same investment takes considerably longer to return.
The benefits worth counting include direct labour reduction, lower overtime and weekend working at shipping peaks, higher utilisation from reduced downtime, fewer quality losses from bad seals and mislabelled cartons, and lower injury risk from repetitive heavy lifting. Maintenance also becomes more predictable, since robotic palletizers have a mean time between failures of roughly 80,000 to 100,000 hours, which makes planned maintenance straightforward.
There is a cost that must not be left out. An integrated line raises the skill requirement for maintenance staff, so budget for training and, where necessary, for additional maintenance headcount. Approving a project purely on “automation removes operators” arithmetic tends to end with a line nobody can keep running.
Why shipping and packaging labour reduction is urgent in Thailand
The shortage now reaches factory operator level
Research by the Japan External Trade Organization found that 40.4% of Japanese companies operating in Thailand face a shortage of people, and 42.3% report the shortage as serious even at factory operator level. The old assumption that labour is easy to secure in Thailand no longer holds.
The pattern is global. Research by PMMI, cited by TAWI, reports that 95% of packaging operations struggle to find skilled operators and technicians. Packaging is often treated as unskilled work, but changeover and fault recovery require real experience, and that experience is exactly what cannot be hired.
Where wages and policy are heading
Wages tell the same story. The January 2024 minimum wage revision set rates between 330 and 370 baht, an average increase of 2.4%. Wage growth of 3.8% in 2023 was the lowest among the nine ASEAN countries surveyed, which looks calm in isolation. But the current government has pledged to raise the daily minimum wage to 600 baht by 2027. Whether that target is met in full is uncertain; planning a payback period on the assumption that today’s wage level persists is not prudent. Companies are already responding, with 27.9% having started automation and 27.5% planning to do so.
Lifting existing plants while new factories decline
There is also a wider signal. Thailand’s Ministry of Industry recorded 59 new factory registrations in January 2026, down 59.9% year on year and the twelfth consecutive month of decline. When new capacity stops arriving, output growth has to come from improving existing plants. Shipping and packaging is the most legible place to look, because the back end of the plant can be re-engineered without touching product specification or revalidating quality.
A staged roadmap
Building a fully integrated line in one step is realistic neither for the budget nor for the shop floor. Four stages work better.
| Stage | Typical duration | Activity | Output of the stage |
|---|---|---|---|
| 1. Baseline | 1 to 2 months | Measure throughput, downtime breakdown and headcount per station, list every format | Data and an identified true bottleneck |
| 2. Whole-line design | 2 to 3 months | Set target capacity, compare layouts, define integration scope and phasing, estimate budget | A design and an investment plan |
| 3. Partial deployment | 3 to 6 months | Install the bottleneck station plus the conveying and buffering around it | Measured proof of the effect |
| 4. Completion | 6 to 12 months | Automate remaining stations, integrate control, connect host systems, train and standardise | An integrated line and an operating standard |
Skipping stage 2 and jumping straight to deployment is what causes rework, because conveyor heights and control architecture will not match when the next machine arrives. Design fees are small against total investment, and cutting them is the least profitable saving available.
Frequently asked questions about packaging line automation
How much does packaging process automation cost
No single benchmark applies, because cost varies with format count, required capacity and how the line connects to existing equipment. The structure, however, is always the same four layers of equipment, transport and integration, installation, and commissioning. Budgeting from machine list prices alone leaves out the transport, integration and installation layers, which is why early estimates and final totals diverge so widely. Set an allowance against each of the four layers before issuing the request for proposal.
Should a packaging line layout be straight or U-shaped
It depends on the shape of the shipping area and on how many people you want to run the line. A straight run is simplest where a long clear space exists and infeed and outfeed can sit apart. A U-shape is better when a small team must supervise several stations or when infeed and outfeed need to be close. An L-shape suits sites where columns or existing equipment must be worked around. In every case, fix the buffer lengths between stations before calculating total line length.
Which station should shipping and packaging labour reduction start with
Start with the station that stops the longest, not the one with the most people. Automating a station that runs smoothly with several operators will not speed up a line that still accumulates product before and after it. Take one week of downtime data and find the station where changeover plus jams add up to the most minutes. Very often it is sealing or labelling, which need few people but stop frequently.
Is it better to start fully automatic or semi-automatic
For a single-shift plant with frequently changing formats, semi-automatic is the more rational start. Investment and installation time are lower, so real operating data can inform the next decision. For a plant running two or more shifts with routine overtime and stable formats, designing a fully automated integrated line from the outset returns the investment faster.
At what point should an integrator be involved
As soon as the baseline stage is complete, and in any case before whole-line design begins. Approaching an integrator after the machine model has been chosen limits them to proposing a layout around that model. Bringing measured data, a target capacity and the site constraints, with no equipment preselected, produces far better proposals.
Conclusion
Packaging line automation is decided by whether the chain from case erecting to the outbound conveyor is designed as one system, not by which machine is purchased. Installing single machines in isolation leaves people between the machines, lets stoppages propagate, fragments the data and makes changeover dependent on individuals, which is how an executed investment produces no measurable result.
The sequence that works is to measure throughput and downtime by station and identify the true bottleneck, then design the whole line and split the investment into phases. Write every format condition and a real acceptance criterion into the request for proposal, and compare quotations on whether integration and commissioning hours are included rather than on headline totals. Read cost in four layers, and count utilisation, quality loss and injury risk alongside labour in the payback.
In Thailand the shortage of people now reaches factory operator level, new factory registrations are falling, and growth has to come from existing sites. Shipping and packaging is the area where the return is easiest to read, because it can be changed without touching product specification. Design the whole, invest in stages.
TOMAS TECH supports Japanese manufacturers in Thailand and across ASEAN with factory automation and robot system integration alongside production management systems, including the integrated design of packaging lines. We are happy to help at the early stage as well, whether that means setting up downtime measurement or making competing quotations comparable. Start by telling us about the situation on your floor through our contact page.
References
- Packaging Robots Market Analysis — Growth forecast for the packaging robot market from 2025 to 2030, with a compound annual growth rate of 10.9%, a market opportunity increase of USD 5.59 billion, and Asia Pacific driving 41.2% of that growth.
- Case Packing and Palletizing Automation in 2026 — Analysis of case packing and palletizing automation adapting to stock keeping unit growth and tighter labour markets, including the 105% year over year rise in North American robot orders, mean time between failures of 80,000 to 100,000 hours, and end-of-line payback in 12 to 24 months.
- Packaging Industry Trends 2026 — Packaging industry outlook for 2026, reporting that 95% of packaging operations struggle to find skilled operators and technicians, and comparing semi-automatic with fully automatic solutions.
- Labour Shortages in Thailand and the Outlook for Minimum Wage — Report by the Japan External Trade Organization on labour shortages at Japanese companies in Thailand and on minimum wage trends, including the 40.4% facing shortages and the 27.9% that have already started automation. Published in Japanese.
- Thailand Records Sharp Fall in New Factory Registrations in January — News report based on Thai Ministry of Industry figures, covering the 59 new factory registrations in January 2026, a fall of 59.9% year on year and the twelfth consecutive month of decline. Published in Japanese.
