Pre-Delivery Repeatability Positioning Accuracy Testing: Standards, Methods, and Acceptance Criteria

Pre-delivery repeatability positioning accuracy testing verifies that machine tools, industrial robots, and precision motion systems can consistently return to commanded positions before shipment, ensuring compliance with international standards such as ISO 230-2 and ISO 9283. This article explains the testing principles, sample types, measured parameters, standard test procedures, statistical evaluation methods, and typical acceptance criteria applied during factory acceptance inspection. Whether you are a manufacturer preparing equipment for delivery or a buyer defining acceptance requirements, this guide provides the technical foundation needed to understand the inspection process.

Technical Overview

Repeatability positioning accuracy is one of the most fundamental performance indicators of any numerically controlled machine or robotic system. It quantifies how closely a motion axis or end-effector can return to the same commanded position over a series of repeated approaches under identical conditions. Unlike absolute positioning accuracy, which describes the deviation between the commanded position and the mean attained position, repeatability describes the dispersion or scatter of attained positions and directly reflects the mechanical quality of drives, guideways, bearings, and control compensation settings.

Pre-delivery repeatability positioning accuracy inspection is the final verification performed by manufacturers, or by accredited third-party laboratories on their behalf, before equipment leaves the production facility. The test confirms that assembly, alignment, drive tuning, pitch-error compensation, and backlash compensation have been correctly implemented, and that the machine will meet the specifications promised in the purchase contract. The documented results of this factory acceptance test form the contractual baseline for subsequent installation, commissioning, and warranty assessment at the customer site.

The internationally recognized test codes are ISO 230-2 (Test code for machine tools, Part 2: Determination of accuracy and repeatability of positioning of numerically controlled axes) and its regional adoptions EN ISO 230-2 and GB/T 17421.2. Industrial robots are evaluated under ISO 9283 (Manipulating industrial robots, Performance criteria and related test methods), adopted in China as GB/T 12642, while coordinate measuring machines follow the ISO 10360 series. These standards define target layouts, approach directions, cycle counts, statistical calculations, and reporting requirements so that results from different manufacturers and laboratories remain directly comparable.

Testing Samples

CNC Machine Tools

Vertical and horizontal machining centers, CNC lathes and turning centers, five-axis machining centers, surface and profile grinding machines, wire-cut and die-sinking EDM machines, and CNC drilling and tapping centers. Both linear axes and rotary axes are inspected individually.

Industrial Robots and Collaborative Robots

Six-axis articulated robots, SCARA robots, delta parallel robots, gantry robots, collaborative robots, and rail-mounted manipulators. Pose repeatability of the tool center point (TCP) is verified at standardized test poses distributed within the working envelope.

Linear Motion Systems and Modules

Ball screw driven linear stages, linear motor stages, multi-axis Cartesian modules, belt-driven actuators, and precision tool changers delivered as standalone units or as integrated subsystems of larger machines.

Rotary Tables and Indexing Mechanisms

Direct-drive and worm-gear rotary tables, fourth and fifth axes, indexing heads, drum turrets, and pallet changers, where angular positioning repeatability directly determines part quality and cycle consistency.

Precision Measurement and Special-Purpose Equipment

Coordinate measuring machines, vision inspection systems, dispensing and bonding machines, semiconductor wafer handling and packaging equipment, and automated assembly stations with positioning-critical operations.

Testing Items

A complete pre-delivery inspection reports the following quantities for each tested axis, calculated separately for each approach direction in accordance with ISO 230-2:

• Unidirectional repeatability of positioning at each target position, expressed as a band of four standard deviations covering approximately 95% of attained positions.
• Axis repeatability of positioning as the maximum value found among all target positions.
• Mean unidirectional positional deviation at each target and the range of systematic positional deviation over the axis travel.
• Positioning accuracy, defined as the width of the band containing all mean positional deviations expanded by two standard deviations at each end.
• Reversal value, commonly called backlash, at each target and the maximum axis reversal value.
• Bidirectional repeatability of positioning, combining the scatter of both approach directions with the reversal value.
• Angular repeatability and positioning accuracy of rotary axes, expressed in arc-seconds.
• Drift of positional deviation as a function of warm-up and thermal state, where required by the specification.
• For robots per ISO 9283: pose repeatability, pose accuracy, distance accuracy and repeatability, and where specified, overshoot and settling time.

Testing Methods

Environmental and Machine Preparation

All measurements are performed in a thermally stabilized inspection area at the standard reference temperature of 20 °C defined by ISO 1, typically maintained within ±1 °C to ±2 °C. The machine is levelled, anchored where required, and exercised with a warm-up program so that drives, guideways, and the spindle reach thermal equilibrium. Air temperature, pressure, and humidity sensors feed the wavelength compensation of the measuring instruments, and the thermal expansion coefficient of the material under test is entered into the compensation unit, conventionally 11.5 ppm/°C for steel and adjusted for cast iron or aluminum where applicable. External vibration and draughts are excluded from the measurement area.

Laser Interferometer Procedure for Linear Axes

A frequency-stabilized He-Ne laser interferometer with sub-micron resolution and compensated accuracy better than ±0.5 ppm serves as the reference instrument. The laser head is fixed to the machine structure, a retroreflector is mounted on the moving carriage, and the beam is aligned parallel to the axis travel within the tolerances given in ISO 230-2. At least five target positions are selected and evenly distributed over the useful travel, positioned no closer than approximately 5% of the travel length from each end stop. Starting from a park position, the axis approaches each target by rapid traverse with a short dwell of typically 3 to 5 seconds, and the sequence is repeated for the required number of cycles, five being the minimum. The entire procedure is then repeated with approach from the opposite direction to obtain bidirectional data. Dead-path error is minimized by setting the interferometer reference at the first target position.

Rotary Axis Testing

Rotary axes are measured with a polygon mirror and autocollimator, with laser interferometer angular optics, or with a precision rotary axis calibrator, depending on the required measurement uncertainty. Target angles are distributed around 360 degrees, and bidirectional cycles are executed analogously to linear axes. Results are reported in arc-seconds and evaluated with the same statistical framework.

Robot Pose Repeatability Testing per ISO 9283

The robot is programmed to move to test poses located at the corners, edge midpoints, and center of a standard test cube inside the working envelope, using defined velocities, payloads, and approach directions. The attained tool center point coordinates are recorded for at least 30 repetitions at each pose, typically with a laser tracker or a high-accuracy contact measurement system. Pose repeatability is calculated as the radius of a sphere derived from the mean distance of the attained poses from their barycenter plus three standard deviations, and is expressed in millimeters.

Data Reduction and Statistical Evaluation

For each target position and direction, the mean positional deviation and the estimator of the standard deviation are computed from the individual positional deviations. Unidirectional repeatability equals four standard deviations, while bidirectional repeatability combines the scatter of both directions with the reversal value. Positioning accuracy is evaluated as the difference between the largest and smallest values obtained when two standard deviations are added to and subtracted from the mean positional deviation at every target. In accordance with good metrological practice, the measurement uncertainty of the instrument chain must not exceed one third, and preferably one fifth, of the tolerance being verified.

Compensation and Re-Verification

If the first measurement run exceeds the specified limits, pitch-error compensation tables and backlash compensation values are updated in the CNC or motion controller, and mechanical adjustments such as bearing pre-load or coupling tightening are performed where necessary. The full test sequence is then repeated under final machine conditions, and only a clean re-run is documented as the pre-delivery result.

Documentation and Reporting

The test report identifies the machine model and serial number, the axes tested, the measuring equipment with its calibration status, environmental conditions, test parameters, individual readings and statistical results, a statement of measurement uncertainty, the applicable standards, and a clear pass or fail conclusion against the contracted acceptance values. Reports are countersigned by manufacturer and customer representatives whenever the factory acceptance test is witnessed.

Applications

Pre-delivery repeatability positioning accuracy inspection is applied wherever consistent positioning determines product quality, yield, throughput, or safety:

• Aerospace component machining, where tight-tolerance pockets, holes, and five-axis surfaces demand verified axis behavior over long travels.
• Automotive powertrain manufacturing, including engine blocks, transmission housings, and high-volume transfer lines requiring stable bidirectional positioning.
• Semiconductor and electronics equipment, where wafer handling, die bonding, and inspection stages rely on micron-level repeatability.
• Medical device production, for validated machining cells and cleanroom automation where process capability depends directly on positioning scatter.
• Precision die and mold making, where finishing operations in hardened steel expose even small axis reversal errors on contour quality.
• Robot-based automation, including welding, dispensing, assembly, and machine tending cells delivered with certified pose repeatability data.
• Contractual factory acceptance testing and witnessed inspections by accredited third-party laboratories for exported machines and public procurement projects.

FAQ

What is the difference between positioning accuracy and repeatability?

Repeatability describes how tightly a machine returns to the same position when the same command is repeated, and is therefore a measure of scatter. Positioning accuracy describes how close the mean attained position is to the commanded position across the entire axis travel. A machine can be highly repeatable yet inaccurate if its compensation tables are wrong, because accuracy errors can usually be corrected by compensation, while poor repeatability indicates mechanical defects that compensation cannot fully cure.

Which standards govern the test?

Linear and rotary axes of machine tools are tested per ISO 230-2, adopted regionally as EN ISO 230-2 and in China as GB/T 17421.2. Industrial robot pose repeatability is tested per ISO 9283 and GB/T 12642, and coordinate measuring machines follow the ISO 10360 series. Purchase contracts may add manufacturer acceptance criteria that are stricter than the standards.

How many positions and repetitions are measured?

ISO 230-2 specifies a minimum of five target positions distributed over the useful travel and five measurement cycles in each direction, producing at least 25 readings per direction. Critical axes and final acceptance tests frequently use more positions and more cycles to reduce statistical uncertainty. Robot pose repeatability per ISO 9283 is normally based on 30 repetitions per pose.

What instruments are used and how accurate are they?

Frequency-stabilized He-Ne laser interferometers with automatic environmental compensation, offering compensated accuracy better than ±0.5 ppm, serve as the primary standard for linear axes. Autocollimators with polygon mirrors or laser angular optics are used for rotary axes, and laser trackers with uncertainty below 0.01 mm are used for robot TCP measurement. All instruments must hold valid calibration traceable to national metrology institutes.

What are typical acceptance values?

Acceptance values depend on machine size, class, and contract. As a general orientation, standard machining centers commonly guarantee repeatability between 0.003 mm and 0.006 mm, precision jig borers and grinders below 0.002 mm, and industrial robots between 0.02 mm and 0.1 mm depending on payload and reach. The contractual value stated in the purchase specification always prevails.

Why must the test be performed before delivery rather than after installation?

Pre-delivery testing verifies the machine in a controlled thermal environment with calibrated instruments, isolates the intrinsic quality of the machine from foundation and site effects, and allows defects or compensation errors to be corrected while the machine is still in the factory. The documented results then serve as the reference baseline for installation, periodic verification, and warranty assessments.

How long does the inspection take?

A single-axis bidirectional test per ISO 230-2, including warm-up, setup, and compensation iteration, typically requires half a day. A complete three-axis machining center with a rotary axis and full reporting usually takes one to two days, while robot repeatability testing per ISO 9283 requires a comparable effort depending on the number of test poses.