Industry Information

Vision Inspection Machine Acceptance: FAT, SAT and Sign-Off

From sample approval to FAT/SAT sign-off - the acceptance stages for a vision inspection machine, what each stage verifies, and the documentation to demand before it is yours.

hzxxs02-dual-table-visual-inspection-machine

The Gap Between Ordering and Sign-Off

The buying process for a vision inspection machine has a logistics half and a verification half. The logistics half - lead time, packing, shipping, export documents, warranty - is covered in Buying an Inspection Machine from China: What to Expect (/posts/buying-inspection-machine-from-china-what-to-expect). The verification half is this article: at each stage between order and sign-off, what do you as the buyer actually check, and what records should exist when the machine becomes yours. The two halves share one principle: whatever was proven before the order - above all the sample test - is the baseline the delivered machine must match. Verification means comparing against that baseline in writing, not accepting "a machine like the one you tested."

Stage 1: Locking the Baseline (Sample Approval)

The sample test is the baseline-setter: your parts run through the machine, the factory returns the images and OK/NG decision log, and the configuration is confirmed to your quality standard. The full walkthrough is in the sample test guide (/posts/vision-inspection-sample-test). For acceptance purposes, two things matter at this stage:
• Get the tested configuration in writing: which camera count, which feeder, which size band and which defect rules were run - this document is what you will verify the delivered machine against.
• Make the sample package mirror production variation: more than one lot of good parts, your worst-case defective parts, and borderline pieces. A baseline built on the supplier's demo parts proves little at sign-off.

Stage 2: Verifying the Delivered Machine

When the machine arrives, the verification question is narrow: does what was delivered match what was tested? Check against the written baseline from Stage 1:
• Configuration match: the camera count, feeder type, size band and defect rules match the documented sample-test configuration - not "a similar configuration with minor differences."
• Documents on hand: ask that the configuration documents and a copy of your sample-run results travel with the machine, along with the spares and tools your quotation listed - make that request part of the order. (What typically ships in the crate is the buying-from-China article's ground.)
• Physical condition on arrival: no shipping damage to the frame, turntable, cameras or fixtures - note any damage before commissioning starts, while responsibility is still with the freight leg.

Stage 3: FAT and SAT - Proving It Runs Where It Will Work

Factory Acceptance Test (FAT) and Site Acceptance Test (SAT) are the standard purchasing stages for verifying a machine before final sign-off. The split:
• FAT (at the supplier's facility, before shipping): the machine runs your parts with the sample-confirmed configuration - the same defect families and dimensions from your sample test. This proves the build. Confirm beforehand whether the FAT runs on the same parts you shipped for the sample test or a fresh lot you send - either is valid as long as the criteria are the same.
• SAT (at your factory, after installation): the installed machine runs on your floor, with your power, your feeding setup and your production environment. This proves that shipping and installation did not change the behavior the FAT validated.
• Both stages should use the same pass/fail references: your documented defect families and dimensions from Stage 1. A test with new criteria invented at FAT time verifies nothing against your baseline.
One bridge between the two articles: the buying-from-China article describes the pre-shipment dry run on your samples - FAT is how a purchasing team turns that run into a formal pass/fail gate. This much is published process: a pre-shipment run on your parts, and commissioning offered remote or on-site. Which acceptance steps are formalized on paper, and whether an engineer attends your site, is stated in the quotation - ask for it explicitly, because suppliers differ in whether they formalize FAT/SAT.

The Written-Record Checklist

Acceptance is only as strong as the records behind it. Before sign-off, these five documents should exist:
1. The tested configuration from the sample run - camera count, feeder, size band, defect rules - matching the delivered machine. 2. The sample-test results - images and OK/NG decision log from your parts - attached to the acceptance record, not referenced verbally. 3. The acceptance criteria - which dimensions and which defect families were demonstrated, at what pass/fail behavior. A usable structure: a seeded set of your known-defective parts, one or more known-good lots, a stated false-reject tolerance, and the OK/NG record logging each decision - the same elements as defining defects for OK/NG inspection (/posts/defining-defects-ai-visual-inspection). 4. The FAT and SAT records - what was run at each stage and what the results were. 5. The installation and training scope - what commissioning covered, and who from your side signed it off. (Commissioning options - remote vs on-site - are compared in the buying-from-China article.)

Frequently Asked Questions

What is the difference between FAT and SAT?

FAT (Factory Acceptance Test) runs at the supplier's facility before shipping; SAT (Site Acceptance Test) runs at your factory after installation. Both verify the same baseline - the SAT proves that shipping and installation did not change the behavior the FAT validated.

Who provides the acceptance documents?

The supplier produces the test records per stage; you hold the baseline documents (tested configuration, sample-test results) from Stage 1. Acceptance works when the two sides match - confirm at order time that you will receive both.

Can our customer or a third-party inspector witness the acceptance test?

Access policies differ between suppliers. Whether your customer's auditor or a third-party agency can witness the FAT or SAT, and what documentation they receive, is a question to raise when you request the quotation - worth raising, because many auditors ask for witnessed acceptance records.

What happens if the machine fails acceptance?

The failed items are corrected and the failed stage is re-run against the same criteria. What a failure does to delivery timeline and payment is defined by your contract - which is exactly why the acceptance criteria belong in writing before the order, not in a discussion on delivery day.

Do we still need the sample test if we plan FAT/SAT?

Yes - they are different jobs: the sample test proves the configuration on your parts before you order; FAT/SAT prove the delivered machine still matches that proof. Without the Stage-1 baseline, FAT/SAT have nothing to verify against.

Next Step

Build your acceptance checklist now - baseline documents from the sample test, configuration-match items, acceptance criteria and the FAT/SAT record list - and raise it when you request the quotation checklist (/posts/vision-inspection-quote-checklist). The logistics side of the same journey is in Buying an Inspection Machine from China (/posts/buying-inspection-machine-from-china-what-to-expect); the incoming-vs-in-line decision shapes where your SAT happens. Contact the team (/contact-us) - Monica (+86 138 5726 5705) or Arya (+86 173 2430 9735), or email monica@aivisioncheck.com.

Read next

More from the inspection knowledge base

More machine vision guides and product spotlights from our engineering team.