FAT vs. SAT for Packaging Equipment: Defining Acceptance Criteria

Compare factory and site acceptance tests for packaging machines, and learn how to write acceptance criteria that both buyer and supplier can verify.

Illustration of a packaging machine on a trial floor with floor markings while an engineer seen from behind observes a test run
Illustration
Short answer

A factory acceptance test (FAT) checks the machine at the supplier's site with the supplier's utilities and your samples, while a site acceptance test (SAT) checks it in your plant on your line under production conditions. Both need written criteria agreed before the test, covering what is tested, how it is measured and who signs.

Applies to: Covers acceptance testing of new standalone packaging machines and small cells bought from a supplier. It does not set test durations, sample quantities or pass limits, and it is not a validation protocol for regulated products.

Many packaging machine disputes start after the machine arrives. The buyer says it does not run as expected; the supplier says it met the order. Often the real cause is that nobody wrote down what “acceptable” meant before the tests began.

The factory acceptance test (FAT) and the site acceptance test (SAT) do different jobs, and each needs criteria that both sides can verify. The sections below show what each test is for and how to write those criteria. Durations, sample counts and numeric limits are left out, because they depend on your product, package, risk and contract. Request the acceptance method in the RFQ as well; the RFQ checklist shows where it fits.

How FAT and SAT fit in a project

A packaging machine project passes through several stages, and acceptance tests sit at two of them. IEC 62381 is a relevant reference for the terminology of FAT, SAT and site integration testing (SIT). Terms vary between suppliers and industries, so define them in your contract rather than assuming everyone reads them the same way.

Timeline from specification and order through design review, build, FAT, ship and install, commissioning, SAT and handover, with notes on where and under what conditions FAT and SAT take place.
Figure 1. Project timeline from specification and order to handover, with FAT and SAT marked. Schematic, not to scale.

The FAT takes place at the supplier’s site, with the supplier’s utilities and your samples and materials. The SAT takes place at your site, on your line, under production conditions. Commissioning sits between them: the supplier or a contractor installs the machine, connects utilities and brings it to a state where it can be tested. Commissioning is not acceptance.

What FAT is for

A FAT lets you find problems while the machine is still in the supplier’s hands and can be modified without shipping, rigging or lost production. It typically covers checks of the build against the specification, review of documents, functional checks of controls and safety functions, and trial runs with your samples. Finding a fault here costs less than finding it after installation.

Its limit is that the conditions are not yours. The supplier’s compressed air, electrical supply, operators and ambient conditions differ from your plant, and the upstream and downstream machines are absent or simulated.

What SAT is for

A SAT shows that the installed machine works in your plant: with your utilities, connected to your line, operated by your people, and running your production material. It confirms that transport and installation did not damage anything and that the interfaces work. Interface problems are the main reason a machine that passed FAT does not perform in the line, which is why the packaging line integration checklist should be completed before the SAT plan is fixed.

FAT vs. SAT comparison

The table shows how the two tests differ. It describes typical practice. Your contract may define them differently.

Aspect FAT SAT
Where Supplier’s site Your site
Purpose Confirm the machine meets the specification before shipment; find and fix defects early Confirm the installed machine works in your plant and line, and release it for handover
Samples and materials Your samples and materials sent to the supplier, as agreed; representative of production, including the difficult cases Production material from your normal supply, or material agreed for the test
Operating conditions Supplier’s utilities, workshop environment, supplier’s operators; neighbors simulated Your utilities, ambient conditions, line connections, your operators
Records Test report, deviation list, document review, photos; signed by both parties Test report, deviation list, installation checks, signed handover; kept in your quality or project records
Who is responsible Supplier prepares and runs; buyer witnesses and reviews Joint: supplier, installer and buyer; buyer often owns the site conditions and material supply
If a test fails Supplier corrects, then repeats the failed item at the agreed scope; shipment waits Parties find the cause (machine, installation, material or site), agree a correction and retest; payment milestone may be held

The last row matters most for commercial planning. Decide before the order who corrects a failure, who pays for the retest, and how a failure caused by the site or by your material is treated.

How to write acceptance criteria

An acceptance criterion should be specific enough that two people can run the test separately and reach the same result. Each one needs five elements.

  1. Test object. State exactly what is tested: the machine function, the output, the package quality, a safety function, a document. “Machine works” is not a test object; “machine forms, fills and seals the agreed package format with the agreed product” is.
  2. Duration and sample quantity. Write both in the document and have both parties sign. The right values depend on the product, the risk, the variation you want to see and the process, so this guide proposes none. Agree them with the supplier and, where your quality system requires it, with your quality staff.
  3. Measurement method. Say how each result is measured: which instrument, which method, who reads it, how often. For package checks, such as seal quality or fill weight, refer to the agreed test method and calibrated equipment rather than a visual impression. If you use an inspection device for verification, define how that device itself is checked.
  4. Pass or fail rule. State in advance what result counts as a pass. A rule might be a limit on a measured value, a count of allowed defects, or a function performing as described. Say how out-of-range results are treated: whether one failure fails the whole test, or whether a defined retest is allowed.
  5. Sign-off. Name who signs, in which role, and what the signature means. Define what happens to open items (a punch list) and whether they block payment or handover.

Conditions that must be written down

Add the conditions under which the test runs. These are the common sources of argument.

  • Product, package and material used, with batch records
  • Format and settings used, and whether the supplier may adjust settings during the test
  • Utilities supplied and their measured state
  • Operators and their training status
  • Which stops, jams or interventions count against the result and which are excluded
  • How output is calculated (net good packs over the agreed period, not display speed)
  • Which tools and instruments are used, and their calibration status
  • Which documents must be present for the test to start

Every criterion should trace back to a line in the RFQ or purchase specification. If a requirement is in the specification but no test covers it, either add a test or note that it is verified another way, such as document review. If a test has no requirement behind it, ask why it is there. Criteria for inspection and reject systems need particular care: a vision inspection system is verified by how it detects and rejects defined defects, and it needs a defined set of defective samples and a method to confirm rejection.

Handling deviations and changes

Every test finds something. Decide in advance how findings are classified, for example as items that block shipment or handover, items that can be closed later with a date, and items accepted as they are with written agreement. Give each deviation an owner and a due date, and record the closing evidence, not just a tick.

Also agree how changes made during the test are treated. If the supplier adjusts a setting, replaces a part or changes software to pass an item, the change should be written down, and the affected tests may need to be repeated. Without this rule, a machine can pass on paper in a state that differs from the one shipped.

For machines that must exchange signals with other equipment, plan how the interfaces are tested at the FAT. Simulators or test rigs can check signals, but only the SAT shows the real exchange with the neighboring machines.

FAT and SAT checklist

Use this as a starting list, then adapt it to your machine and contract.

Before the FAT

  • Acceptance plan agreed and signed by both parties
  • Samples and materials delivered to the supplier, labeled and recorded
  • Document set from the supplier received for review (drawings, schematics, manuals, parts list)
  • Attendees named, including quality and maintenance staff if relevant
  • Safety review status known: the machine’s guarding and safety functions come from the supplier’s risk assessment, which qualified people on both sides should review

During and after the FAT

  • Build checked against the specification and drawings
  • Controls, alarms, recipes and safety functions checked as agreed
  • Trial run with the agreed materials and conditions
  • Results measured with the agreed method and recorded
  • Deviations logged with owner and due date
  • Agreed items closed before shipment, or documented as accepted exceptions
  • Both parties sign the report

Before and during the SAT

  • Installation checked: position, anchoring, utilities, guarding, connections
  • Interfaces to upstream and downstream machines checked
  • Production material and operators available
  • Trial run under agreed production conditions
  • Results measured and recorded using the agreed method
  • Training and documentation handover completed
  • Open items listed with owners and dates
  • Handover signed, with spares and software backups received

Regulated environments

If the machine will be used for products in regulated (GxP) environments, such as pharmaceuticals, your quality system may require a documented verification approach. Relevant references include ISPE GAMP 5, which covers the lifecycle and testing of regulated computerized systems, and ASTM E2500, which describes specification and verification for regulated manufacturing equipment. They may change how test documents are written, reviewed and retained. This guide does not replace them; follow your own quality unit’s procedures.

Questions to ask suppliers

Ask these before the order, so the answers appear in the contract.

  • Do you offer FAT, SAT or both, and which are included in the price?
  • What does your FAT cover, and what is left to the SAT?
  • What do you need from us for the FAT: samples, quantities, timing, documents?
  • Can you provide a draft acceptance plan and a sample report format?
  • Who may attend the FAT, and how are records shared if we cannot attend?
  • How do you handle deviations found in the FAT?
  • How is the SAT scheduled, and who supplies utilities, operators and material?
  • What happens if the SAT fails, including retest conditions and responsibility?
  • How are payment milestones tied to FAT and SAT?
  • Which documents are delivered at each stage?

Limits and on-site verification

No FAT can fully reproduce your plant. Product and material variation across batches, operator habits, ambient conditions and line interactions only appear in production, so an accepted machine may still need tuning after handover. Plan for a stabilization period and agree in writing how problems found in it are handled.

Acceptance criteria for safety functions, electrical compliance and regulated use need qualified review. Safety design follows a risk assessment, for which ISO 12100 gives a general method. Do not use an acceptance test to skip or relax any safeguard, and do not alter guards, interlocks or energy isolation for a test. Final values for duration, quantity and limits must come from your own product, process and quality requirements, and from the supplier’s manual.

Total cost depends on how well the machine performs after acceptance, so see the total cost of ownership framework when comparing offers. Keep the acceptance method consistent with what you wrote in the RFQ.

References

  1. IEC 62381:2024 — Automation systems in the process industry — Factory acceptance test (FAT), site acceptance test (SAT), and site integration test (SIT) — IEC
  2. ISO 12100:2010 — Safety of machinery — General principles for design — Risk assessment and risk reduction — ISO
  3. ISPE GAMP 5 Guide: A Risk-Based Approach to Compliant GxP Computerized Systems (Second Edition) — ISPE
  4. ASTM E2500-25 — Standard Guide for Specification, Design, and Verification of Pharmaceutical and Biopharmaceutical Manufacturing Systems and Equipment Science and Risk Based Approach — ASTM International

Update history

  • : First published.