A cartoning machine acceptance test is the documented set of checks that proves a cartoner forms a carton, loads your product into it, closes and codes it, and ejects it at the rate you agreed, using your own product and cartons. It exists to turn a supplier’s claims into evidence you can re-measure later, and to decide, on paper, whether the machine is fit to be paid for and put into production.
That sounds obvious, and it is skipped all the time. The phrase gets stretched to cover anything from a twenty-minute demonstration to a two-day formality.
Here is the pattern behind most failed sign-offs. The test measures what is easy to measure: cartons counted, cycle time with a technician’s hand on the dial, a stack of ideal samples the vendor brought along in a case. None of that predicts how the machine behaves on a humid Monday, with warped blanks, a new operator, and a leaflet feeder meeting your inserts for the first time.
The test is worth only what it can prove twice. Numbers rather than adjectives. A recorded run rather than a demonstration. Hold that one rule and the matrix on this page becomes something you can hand to a supplier and enforce.
You already know a demo flatters the machine, so we will not pretend otherwise. What follows is a ten-item cartoning machine acceptance test matrix, a defensible way to measure qualified output, and a method for handling deviations without either side pretending they did not happen. First we have to separate four stages that buyers routinely confuse.
Want the matrix built around your carton, your product and your target rate? Send our engineers your URS and carton samples, and we will draft the acceptance protocol before you place the order.
Four stages, not one: FAT, installation, SAT and commissioning
Ask ten buyers what a cartoning machine acceptance test covers and you will get ten answers, most of them merging four separate events into one word. Each stage answers a different question, and a pass at one does not carry to the next.
The factory acceptance test, or FAT, runs at the maker’s plant before shipment. Your product, your cartons, your target rate, with the engineers who built the machine standing beside it. It is the cheapest place to find a problem, because a fix here takes hours instead of weeks.
Installation checks come next. The machine has arrived. Confirm it survived the trip, sits level, is wired to the drawings, and is fed by the utilities you specified rather than the ones in the brochure.
The site acceptance test, or SAT, reruns the core checks in your plant, on your compressed air and your power, with the upstream and downstream equipment connected. A machine that passed in a clean factory can still stumble in the environment it will really live in.
Commissioning and training close the sequence, and this is the stage buyers most often leave out of scope. It is where the machine stops being the vendor’s and becomes your team’s. If you run other machines, you will recognize the logic: it mirrors the shared acceptance criteria you apply when signing off an erector, even though the parts and the checks differ.
The vocabulary is borrowed from qualification practice in regulated manufacturing. EU GMP Annex 15, at sections 3.4 to 3.7, describes how supplier testing and a User Requirements Specification (URS) relate. We cite it as a framework reference only. It is not a certification of this machine, and not a claim that your line is compliant with anything. Treat it as a skeleton: define what you need first, then test against that definition.
Fix the boundaries and the samples before anything runs
Run a cartoning machine acceptance test without locked boundaries and you will spend the sign-off arguing about scope instead of watching the machine.
Start with the machine. The unit that ships is the unit you accept, so record the model, the build revision and the serial number rather than accepting “a unit like the demo”. Then list every format part, magazine and carton size the order covers. A second carton size that is merely “available” is not part of acceptance unless it is on the order and on the line on the day. The downstream interface belongs here too: the connected-line handoffs that decide whether closed cartons reach your case packer cleanly are part of the acceptance scope.
Then fix the samples, because they decide how honest the test is. It is common for a first FAT to run only on the product the vendor happened to have in stock, which is flat, clean and friendly. Your product is not always any of those things.
Build a sample set with the easy case and the hard one. On the product side, push the lightest units, the awkward-to-push ones, and the ones most likely to tip. Cartons should span the typical blank, the thinnest and thickest board your spec allows, and the most warped batch you can reasonably expect. Do not forget leaflets: real stock, the thinnest paper you buy, in its most curled state.
A test that runs only on pristine samples proves the machine can run. It does not prove it can run for you.
The cartoning machine acceptance test matrix
Ten checks carry most of the weight in a cartoning machine acceptance test. Run all of them on your product and cartons, and record each result as a number, a time or a photo rather than a tick. These are the items to write into the contract and re-measure at the SAT.
Build and scope
Confirm the delivered machine matches the order. Check the model and revision against the purchase order, walk the general arrangement drawing against the physical unit, and verify the format parts, magazines and tooling against the order list. Note the utilities the machine actually needs: air pressure and consumption, voltage, and floor space with service access. Evidence is a signed build record and the as-built drawings.
Infeed and grouping
A cartoner is only as steady as the product stream reaching it. Test how your product is supplied, grouped and spaced into the infeed at the target rate, including what happens when the supply pauses. Does the machine run empty and waste cartons, or hold? Does grouping stay in step at the top of the rate range? Flat items that shingle need watching here too, and the thin-card singulation specified for carded products is the same problem in miniature. Record the infeed pitch, the grouping pattern, and the behavior on a deliberate supply gap. The evidence is a timed run with the upstream supply in place.
Carton opening
The magazine and the opening station decide whether the line starts cleanly or fights you all day. Load a full magazine of your blanks and confirm the pick releases one carton at a time, that the carton opens square, and that it holds its shape on the pocket or conveyor. Watch for double feeds, missed picks and cartons that open but will not stay open.
Loading and leaflet insertion
Now the product itself. Confirm your product is inserted or pushed into the open carton without damage, at the right orientation and count. If the product carries a leaflet, test the feeder on your real inserts: does it pick a single leaflet, fold it correctly, and place it without a misfeed? Run the worst combination you have, such as the most curled leaflet with the most awkward product. Trigger the missing-product and missing-leaflet detection deliberately.
Closing and coding
The carton has to close the way the design intends, whether that is a tuck flap, a glue seam or a combination, and it has to close on a full carton as cleanly as on an empty one. Run the closure and confirm the flap or seam holds through the handling the carton will see. Then test the coding: date, lot, and any barcode. Verify print quality, position and legibility, and confirm the code matches the batch you are running. A closure that looks fine on an empty carton and pops on a full one is a fail, not a note.
Inspection and reject
Anything the machine claims to detect should be proven within the cartoning machine acceptance test with a real defect, not a simulation. Feed a carton with a missing leaflet, a wrong count, an open flap or an unreadable code, and confirm the detection triggers and the reject mechanism removes that exact carton. Then confirm the reject is definite: the carton lands in the reject bin and does not drift back into the good stream. Test the reject confirmation, the counter and the alarm. This is the check that decides whether a bad carton reaches your customer.
Qualified output
This is the headline number, and the next section covers how to measure it. In the matrix, treat it as a single line: a timed run on your product and cartons, with good cartons counted against cartons consumed, at or above the agreed rate. Report it as good units per minute over a representative run length, not a peak minute.
Stops, restart and recovery
A machine that stops constantly is worse than one that stops for a real reason, because it teaches operators to ignore alarms. Count nuisance stops over a long run, and treat a high number as a defect rather than a quirk. Then test recovery: clear a deliberate jam, restart, and confirm the machine returns to rate. Test the safety interlocks properly too. Confirm each interlock, not just the guard: open the guard and the machine must stop, and the emergency stop must latch and demand a deliberate reset. EN 415-1 and EN ISO 12100 frame the machine-safety expectations in Europe, and ANSI/PMMI B155.1-2023 plays the same role for machinery in North America.
Handoff and changeover
The cartoner is one machine in a chain, so a cartoning machine acceptance test has to look past its own discharge. Confirm the closed cartons leave in the orientation, grouping and pitch your case packer expects, or with a gauge that reproduces its infeed. The signals that let the next stage start and stop in step belong in the acceptance scope as well. Then handle changeover the way it happens in production: hand the job to your own operator, start the clock at the last good carton of the first size, and stop it at the first good carton of the next.
Documents and release
The paperwork is part of the cartoning machine acceptance test. Without it, the first breakdown becomes a guessing game and the next spare part becomes a shipping delay. Collect the general arrangement drawing, the pneumatic and electrical diagrams, the PLC backup and the changeover procedure. Obtain the wear-part list with lead times and confirm the spares kit matches it. Confirm the manuals are in the language your team reads. A supplier working to ISO 9001:2015 should hand this over as routine, and the acceptance run is your moment to insist on it. Release the machine formally, or release it conditionally against a dated list of open items before handover.
Building the matrix is faster with a second set of eyes on your URS. Send us your carton drawings and target rate, and our engineers will turn your requirements into acceptance criteria you can put in front of any supplier.
How to measure qualified output without inventing a pass line
Qualified output decides whether the machine was worth buying in any cartoning machine acceptance test, so it deserves a method rather than an opinion.
Fix the run first. Agree the run length in advance, long enough to expose drift. A single clean minute proves nothing. Half an hour, or a full shift, with the good-carton count divided by the elapsed time, gives a rate you can plan around.
Decide how downtime counts before you start, not after. Agree what stops are attributable to the machine and what stops are not, and record both. If the line stops because the operator ran out of cartons, that is a different event from a stop caused by a mispick, and the two should not be blended into one comfortable average.
Then resist the urge to demand a universal pass line. There is no single correct cartons-per-minute figure, because it depends on carton size, loading style, product and closure method. Horizontal cartoners are commonly quoted somewhere between tens and a few hundred cartons per minute depending on those variables, and those figures are reported industry ranges and benchmarks, not guarantees. The only number that matters is the one your own run produces.
So write the criterion as “agreed net output, measured on buyer’s product and cartons over the agreed run length”, and tie it to your URS and the contract. Your cartoning machine test criteria should be numbers the supplier signed up to, measured the same way twice, at the FAT and again at the SAT.
Deviations, re-tests and the open-item list
A serious acceptance run produces findings. A test with none at all is more often a sign of a shallow checklist than a perfect machine, so plan for deviations instead of being surprised by them.
The contract should say, in advance, how a deviation is recorded, who decides whether it is acceptable, and what happens if it is not. Some findings are cosmetic and can be accepted on the record. Others change whether the machine meets its purpose, and those cannot be waved through. Both need a written home.
Where a finding is fixable, agree the fix and the re-test, and use the same method as the original so the two results are comparable. Where a finding is still open at sign-off, release the machine conditionally and keep a dated open-item list with an owner for each line. Conditional release is normal on real projects. What is not normal is an open-item list that disappears once the invoice is paid, which is why payment should stay tied to closing it.
Where training belongs in a cartoning machine acceptance test
Training is easy to treat as a nice-to-have and a nuisance to schedule. It belongs in a cartoning machine acceptance test, because an untrained team cannot tell you whether the machine is ready.
Make operator competency an acceptance item. Your operators, not the vendor’s technician, should run the changeover, clear a jam, restart after a stop, and call up a recipe on the HMI while someone watches and signs. The vendor can demonstrate. Your team has to perform.
Ask for a written training record: who was trained, on what, and against which procedure. Keep the manuals where operators can reach them, in their language. When training is a line on the acceptance checklist rather than a favor at the end, handover stops depending on whether one engineer is still reachable by phone.
Why a passed FAT still stalls on site
The machine cleared its factory test, and three weeks after installation it is idling between stops. The cause is rarely the machine, and it is almost always one of four things the FAT could not see.
Product or upstream supply tops the list. The FAT ran on a tidy sample; production feeds the machine from a pallet that has been handled badly, or from an upstream station that groups product differently than the test assumed. Utilities trip up plenty of lines too: the factory test ran on clean, steady compressed air, and your plant’s pressure dips whenever another machine cycles. The cartoner notices before anyone else does, and the symptom looks like a sensor fault.
Site conditions matter as well. A machine that passed on a level factory floor can struggle on one that is not, or in a bay with no room to swing a changeover part. And sometimes it is protocol: a handshake signal between two machines that was never tested together, because each passed its own acceptance in isolation.
None of these are machine defects, and all of them are why a cartoning machine acceptance test has to reflect the plant it will live in, not just the product it handles. If you are still weighing what a supplier should answer before you buy, the questions to ask before buying are a useful starting point, and the machine’s own published cartoning machine test criteria are what you turn into the acceptance numbers.
Where the sign-off happens in a cartoning machine acceptance test
It is not a box to tick on the way to delivery. It is the moment a claim becomes evidence.
Keep the four stages separate. Lock the boundaries and choose samples that include your worst board, your most awkward product and your thinnest leaflet. Walk the ten-item matrix and record each result as a number, a time or a photo. Measure qualified output over a run long enough to mean something, and write every one of those numbers into the contract so they can be checked again. Handle deviations on the record, and let training and documents count as real acceptance items.
Do that, and the machine you accept is the machine you run. Skip it, and you will be the one on the phone three weeks later, asking why a brand new cartoner cannot do what the report said it could.
Ready to turn your requirements into acceptance criteria? Send us your URS, carton drawings and target rate. Our engineers will draft a cartoning machine acceptance test you can hold any supplier to, backed by a quote that stands behind it.
