You fly in with tickets booked and hotels arranged, and the machine is nowhere near ready to ship. A compromise gets negotiated, the equipment is released, and start-up is plagued with in-plant modifications and delays.
That scenario plays out because the buyer planned the FAT as a list of things to watch. The decision that actually falls to you on the day is whether a machine running below its quoted rate is a signature or a punch-list line.
That decision is settled by the numbers written into the protocol before the build slot, not by what anyone observes in the workshop. A factory acceptance test (FAT) is the witnessed run at the builder’s plant, against a protocol both sides signed, before the machine ships and before the balance milestone is released. Equipment selection is already finished.
The FAT protocol is a signed declaration confirming that the machine conforms to the contract’s technical conditions. The beverage machinery industry’s model form (ABMI, circa 2009) carries the equipment designation, serial number, test date, test location, and an annex for attendees and deviation notes.
Write the protocol before a test date is set. Contracts typically require seven days’ notice before testing begins, and some specify “no earlier than two weeks from the date when the notification is received by BUYER.”
Name every test, its acceptance criterion, who witnesses, and who signs. For a packaging machine, the test list covers safety circuits, rate, dose or fill-weight accuracy, seal integrity, changeover, reject handling, and documentation.
Those numbers come from the targets you fixed at equipment selection: packs per minute, pack format, fill weight and tolerance. The rate-test duration, the pack count, the allowable reject percentage, and any machine fill tolerance are numbers your contract and your product govern. Agree them before the build slot.
A protocol that reads “verify performance” with no run duration or tolerance band attached is not enforceable.
The test date sits four to six weeks before the ship date. The builder exercises the machine for several days before the buyer arrives so the speed demonstration means something.
That places your deadline for getting material to the plant weeks before the day you fly.
Ship your own product, your own film reel, and your own cartons in production condition. Send enough quantity for a sustained run. A rate demonstrated on the builder’s demo product is a different number from the same rate on your product at production moisture, particle size, temperature, and film.
Machinery powered up for the first time fails more often than it runs. Mechanical and electrical failures are common in the first hours.
The builder’s pre-test exercise window is where those get cleared before you arrive. Greg Jacob, who runs acceptance tests on the builder side at Allpax Products, puts in-plant corrections at three to four times the shop time.
At minimum, the lead project engineer, project manager, and maintenance manager should attend. If the FAT will double as first operator training, bring operators and technicians as well.
On a multi-lane sachet machine, dose accuracy over a counted sample is what the FAT settles. Confirm the builder will run your product on your film and write the agreed rate, weight tolerance, and reject percentage into the protocol before you fly.
The test sequence runs in the order the machine has to prove itself. Dry cycle and safety-circuit verification come first, then product.
The weight test has a published anchor most protocols never cite. NIST Handbook 133 (2026 edition) Table 2-1 sets the sample size by inspection lot: 12 packs for a lot of 12 to 250, with zero allowed to exceed the Maximum Allowable Variation (MAV). Above a 3,200 lot, 48 packs are sampled and one may exceed the MAV.
Table 2-5 sets the MAV by labeled quantity: 10% below 36 g, so a 5 g stick pack’s MAV is 0.5 g. For packs labeled above 117 g up to 154 g, the MAV is 9.0 g.
Your FAT weight criterion has to sit tighter than that rule, because that rule is what the finished packs get measured against in commerce.
Seal acceptance names a method, not an adjective. ASTM F88 measures peel force on a 1-inch strip in three technique variants. Seal strength and seal integrity are separate questions: strength measures force to separate, integrity measures whether the seal is free of channels and microleaks.
The documentation review checks legal objects. The EU Declaration of Conformity rests on Machinery Directive 2006/42/EC, in force today.
From 20 January 2027, Machinery Regulation (EU) 2023/1230 applies. Only certificates from notified bodies designated by EU countries carry legal value.
A remote or video-witnessed FAT can verify that the machine exists, cycles, and produces packs. What it cannot verify is anything the witness would choose to look at, because the controls and the camera both stay with the builder.
The practice that settled during the travel-restriction period depended on three preconditions: both parties recording, the protocol pre-approved before the session, and a signature log documenting who witnessed what. For a first-of-type machine the missing element is not image quality but the ability to direct what gets examined.
FAT results fall into three states: accepted, accepted with comments (the punch list), or rejected.
A deviation is blocking if it involves a safety function, a performance shortfall against a criterion the protocol named, or a missing conformity or calibration document. The machine stays in the shop. The builder corrects it, you retest, and the retest result is recorded.
A non-blocking deviation joins the punch list and travels with the machine. It closes at installation, with a named owner and a target date.
Anything corrected on the spot gets retested before you leave. Record the correction and the retest on the same sheet.
A FAT proves the machine under the builder’s utilities, product handling, and ambient conditions. It cannot prove yours. Equipment can pass in the shop and fail during site commissioning because of utility fluctuations, shipping damage, or ambient differences.
Site acceptance testing after installation, commissioning under your conditions, and pharmaceutical IQ/OQ/PQ validation are separate stages with separate document sets.
The buyer-signed FAT certificate is the shipment gate. Contract language is explicit: “the SELLER may not ship the Equipment until the Factory Acceptance Certificate has been signed by BUYER.”
Full acceptance means all criteria met, certificate signed, machine ships.
Conditional acceptance means open items are recorded on the punch list and carried to installation. The machine ships with a list of what remains, and each item has a named owner.
Not signing is not a way to hold the shipment gate shut. Under the beverage machinery industry’s model protocol (circa 2009), a customer who fails to counter-sign within seven days of the test date forfeits the gate.
If that failure is not attributable to a material breach by the supplier, the certificate is deemed validly issued on the supplier’s single signature. Shipment proceeds. Silence can be acceptance.
The FAT report you take home must contain the signed protocol, the test results against each criterion, the attendee list, the deviation notes, and the certificate itself. That report is what you defend the machine with when something surfaces at startup.
Before the date is fixed, get three things in writing. Name the test list with quantified pass/fail criteria, the run duration and pack count for the rate test, and the date your product and film must arrive at the builder’s plant.
The test date sits weeks before the ship date, and the machine is exercised before you arrive. Your strongest position is the protocol and the calendar, not the workshop. Every open item that travels with the machine needs a named owner and a close-by date at installation, because that is where the FAT stops and the site acceptance test begins.
