Once a CNC machine leaves the factory, correcting a configuration error becomes slower and more expensive. A factory acceptance test, or FAT, gives the overseas buyer a controlled opportunity to verify the approved configuration, machine geometry, cutting performance, safety functions and documentation before shipment release. The FAT is not a ceremonial machine demonstration. It is a written test plan with acceptance limits, evidence and named responsibility for any deviation.

Freeze the Approved Configuration First
The test cannot prove compliance unless the parties agree what was ordered. Prepare a configuration schedule listing working area, travels, spindle model, tool interface, motors, drives, controller, vacuum arrangement, extraction connection, sensors, electrical voltage, software options and included accessories. Record serial numbers where available. If a substitute component is proposed, its technical effect and approval should be documented before testing. A photograph of the nameplate is more reliable than a later recollection of the model.
Geometry and Performance Evidence to Capture
Choose tests that can be repeated after installation. Record ambient conditions, warm-up state, instruments and the exact location of every measurement:
- Axis travel, homing repeatability, squareness and measured backlash or reversal error.
- Spindle runout at the taper or collet and at a defined test-bar distance.
- Flatness or mapped variation across the usable table after the agreed preparation.
- Dimensions and diagonals of a cut test piece using the buyer’s normal material.
- Vacuum holding, extraction capture, drive temperature and alarm history during a sustained cycle.
A useful measurement record states the instrument, measurement point, machine condition and acceptance limit. A number without those four details is difficult to compare, repeat or use in a purchase decision.
Run the FAT as a Controlled Sequence
A sequence prevents warm-up, parameter changes or operator intervention from hiding the origin of a result:
- Verify documents, serial numbers, options and visible workmanship before power-on.
- Record the parameter and controller backup before test adjustments are made.
- Warm the machine using an agreed cycle and then complete geometric checks.
- Run a representative cutting file long enough to expose heat, vacuum and extraction behaviour.
- List every deviation, corrective action, owner and retest result before authorising packing.
Do not change several variables at once. Record the baseline, make one controlled change and repeat the same test. This approach separates a real improvement from a temporary result caused by material, temperature, tooling or operator variation.
Set Acceptance Limits and Release Rules in Advance
Terms such as “high accuracy” or “good finish” cannot control a dispute. Define the tolerance, test length, instrument and material before the FAT. Separate critical failures that block shipment from minor punch-list items that can be closed with evidence. If software or optional accessories are incomplete, specify whether shipment may proceed and what protection remains. The buyer should receive the final parameter backup and test record after all corrections, not the preliminary file captured before tuning.
Demonstrations That Look Impressive but Prove Little
Cutting a large logo in soft material may show that the machine moves, but it does not test diagonal error, small-part holding, repeated tool changes, thermal drift or production cycle stability. A short idle run cannot confirm vacuum performance under leakage or drive temperature during sustained work. Another risk is measuring only one point on the table. The FAT should represent the buyer’s actual process and should include enough duration and geometry to expose the weaknesses that matter after delivery.
Remote Witnessing for International Buyers
When travel is impractical, the buyer can witness the FAT by live video if the evidence is designed correctly. Use a fixed overview camera plus close views of instruments, serial numbers and the controller screen. Show an external clock or uninterrupted sequence to establish continuity. Share the test file and blank report beforehand, then provide original-resolution recordings and measurement photographs. Remote witnessing does not replace objective limits, but it can make a well-structured FAT practical for buyers in Europe, the Gulf, Africa and other export markets.
Documentation That Protects the Investment
The release pack should include the signed configuration, FAT report, deviation register, final backups, electrical drawings, lubrication schedule, consumable list, packing list and photographs before enclosure. Keep original measurement files where instruments export data. These records become the baseline for arrival inspection and commissioning.
Keep the quotation, approved configuration, electrical drawings, parameter backup, inspection results, serial numbers, consumable list and service contacts in one controlled folder. Photographs should show scale and location; videos should include the machine state and test conditions. This evidence shortens remote diagnosis and prevents a later disagreement about what was supplied, measured or changed.
Select the machine size for the intended application, then request a buyer-specific FAT rather than a generic demonstration.
Review the 1500×800 CNC RouterReview the 5600×2100 CNC Router
Can a factory acceptance test be completed entirely by video?
Yes, if objective measurements, uninterrupted evidence, shared test files and clear acceptance limits are used. The buyer should be able to identify instruments, locations, machine state and serial numbers. A promotional video edited by the supplier is not equivalent to a witnessed FAT.
Should the machine be packed immediately after the FAT?
Only after deviations are closed or formally accepted. Photograph the machine condition, loose accessories, corrosion protection, axis restraints and crate before closing. If the machine remains on the floor after acceptance, repeat critical checks or inspect for configuration changes before packing.
