Robot Repeatability Positioning Accuracy Testing: Standards, Methods, and Technical Requirements

Robot repeatability positioning accuracy testing measures how consistently an industrial robot returns to a commanded pose, and it is a critical performance benchmark for manufacturing quality, process stability, and equipment acceptance. Conducted in accordance with ISO 9283 and GB/T 12642, this testing quantifies pose repeatability, pose accuracy, and path performance using precision metrology such as laser trackers. This guide explains the test items, sample types, measurement methods, calculation principles, and typical applications of robot repeatability testing.

Technical Overview

Robot repeatability positioning accuracy testing is a quantitative evaluation of a robot's ability to return to the same commanded position and orientation over multiple cycles. It is one of the most fundamental performance characteristics of industrial robots, directly determining whether the robot can reliably perform tasks such as precision assembly, welding, dispensing, and machine tending.

The internationally recognized test method is defined in ISO 9283 (Manipulating Industrial Robots — Performance Criteria and Related Test Methods), which is adopted in China as GB/T 12642 and in Japan as JIS B 8432. The standard distinguishes between two related but distinct concepts:

• Pose accuracy (AP): the deviation between the commanded pose and the mean of the attained poses. It reflects systematic error caused by kinematic parameter deviations, link tolerances, and calibration quality.
• Pose repeatability (RP): the scatter or dispersion of attained poses around their mean when the robot approaches the same commanded pose repeatedly. It reflects random error caused by joint backlash, servo resolution, mechanical compliance, and thermal drift.

Modern industrial robots are typically highly repeatable but not necessarily highly accurate. Common repeatability specifications range from ±0.02 mm for precision SCARA and small six-axis robots to ±0.1 mm or more for large-payload articulated robots. Verifying these values through standardized testing is essential for robot manufacturers, integrators, and end users, and is frequently performed by accredited third-party laboratories to provide impartial certification data.

Testing Samples

Six-Axis Articulated Industrial Robots

The most commonly tested robot type, covering payloads from a few kilograms to over one ton. All six joints contribute to the overall pose error, so testing is performed at multiple locations within the working envelope to capture worst-case behavior.

SCARA Robots

Selective compliance assembly robot arms used in electronics manufacturing, where repeatability requirements are often the most stringent (±0.01 mm to ±0.02 mm). Testing focuses on the X-Y plane positioning performance.

Delta (Parallel) Robots

High-speed parallel kinematic robots used in food and pharmaceutical picking. Their lightweight structures and high accelerations make dynamic repeatability and stabilization behavior important test parameters.

Collaborative Robots (Cobots)

Lightweight robots with joint torque sensors and compliant drivetrains. Testing verifies whether force-limited joint designs meet the repeatability class claimed by the manufacturer, typically ±0.02 mm to ±0.1 mm.

Cartesian, Gantry, and Multi-Axis Systems

Linear-motion robots, seventh-axis travel units, and integrated cells where system-level repeatability must be verified, including the influence of tracks, positioners, and end-of-arm tooling interfaces.

Testing Items

Depending on the purpose of the evaluation (incoming inspection, type testing, or periodic verification), testing may include the following items defined in ISO 9283 and GB/T 12642:

• Pose repeatability (RP): including positional repeatability (RP_l) and orientational repeatability (RP_a), evaluated as the radius of a sphere containing 99.7% of attained positions.
• Pose accuracy (AP): positional accuracy (AP_p) and orientational accuracy (AP_a) relative to the commanded pose.
• Multi-directional pose accuracy variation (AV): the spread of mean positions when approaching the same point from different directions, revealing backlash effects.
• Distance accuracy (AD) and distance repeatability (RD): the ability to accurately and repeatedly span a commanded distance.
• Path accuracy (AT) and path repeatability (RT): deviation of the actual trajectory from the commanded path and the scatter of repeated traversals.
• Cornering deviations (CR): round-off error at sharp corners, which is critical for gluing, welding, and milling paths.
• Stabilization time and pose overshoot: dynamic settling behavior after reaching a commanded pose.
• Drift tests: pose stability over warm-up time, quantifying thermal influence on positioning performance.

Testing Methods

Test Setup and Measurement Equipment

The standard instrument is a laser tracker with a spherically mounted retroreflector (SMR) or a measuring sphere attached at the robot's tool center point or mechanical interface. Laser interferometers, ballbars, and high-speed camera tracking systems may be used for specific path tests. The measuring device must provide accuracy at least ten times better than the expected robot repeatability, and it must be calibrated and traceable to national measurement standards.

Test Cube and Pose Layout

ISO 9283 specifies a standard test cube located in a plane diagonal to the base within the working envelope. The cube occupies typically 10% of the working volume (up to 80% for path tests). Five reference points are used: a center point P1 and four corner points P2 to P5. The commanded approach directions are parallel to the cube edges so that the robot approaches P2 from four orthogonal directions via P1.

Test Cycle and Repetitions

For pose repeatability, the robot executes a continuous loop: P1 to P2, returning to P1, then P1 to P3, back to P1, P1 to P4, back to P1, P1 to P5, and back to P1. This cycle is repeated 30 times after an initial warm-up cycle, producing 30 measured positions at each target point.

Speed and Load Conditions

Testing is performed at 10%, 50%, and 100% of the rated velocity. Tests are conducted at rated load as the primary condition, and optionally at 0% (no load) and 50% reduced load to characterize load sensitivity. The programmed path style is typically linear interpolation for pose tests.

Data Processing

For each target pose, the mean position of the 30 attained positions is calculated, followed by the standard deviations along the principal axes. Pose repeatability is computed as RP_l = l_mean + 3·S_l, where l_mean is the mean distance from the mean position and S_l is the corresponding standard deviation. This defines a sphere containing 99.7% of the positions under a normal distribution assumption. Orientation repeatability is evaluated analogously from the scatter of angular deviations.

Environmental Control

Testing is conducted in a controlled environment, typically at 20 °C ± 2 °C with temperature gradients and ambient vibration minimized. Load mass, center of gravity, and mounting conditions must be documented, since ISO 9283 results are only comparable when test conditions are identical.

Applications

Robot repeatability positioning accuracy testing serves multiple purposes across the robotics industry:

• Factory acceptance and incoming inspection: end users verify that delivered robots meet the datasheet repeatability class before commissioning.
• Type testing and product certification: robot manufacturers use standardized test results to substantiate published performance specifications, often with verification by accredited third-party laboratories.
• Predictive maintenance and condition monitoring: periodic repeatability measurements detect gear wear, backlash growth, and servo degradation before they cause production defects.
• Process qualification: welding, dispensing, fastening, and precision assembly processes require verified robot positioning performance as part of process capability studies.
• Calibration and compensation validation: after kinematic identification and error compensation, testing confirms the achieved improvement in accuracy and repeatability.
• Quality disputes and benchmarking: standardized test reports provide an objective basis for comparing robots from different suppliers.

Industries that rely heavily on this testing include automotive body shops, electronics and semiconductor manufacturing, aerospace assembly and drilling, medical device production, and logistics automation.

FAQ

What is the difference between accuracy and repeatability in robots?

Accuracy describes how close the robot's average attained position is to the commanded position, while repeatability describes how tightly the robot clusters around its own average position across repeated attempts. A robot can be very repeatable yet inaccurate; accuracy errors can often be corrected through calibration, whereas poor repeatability usually indicates mechanical issues.

Which standards govern repeatability testing?

ISO 9283 is the primary international standard, adopted nationally as GB/T 12642 in China and JIS B 8432 in Japan. ANSI/RIA R15.05-series standards define related point-to-point and path performance evaluation methods for industrial robots in North America.

How many measurement cycles are required?

ISO 9283 specifies 30 valid measurement cycles per test pose for pose repeatability and accuracy testing, performed after at least one warm-up cycle, with measurements taken only after the robot has stabilized at each pose.

What instruments are used for the measurement?

Laser trackers are the most common instruments due to their large measurement volume and micron-level accuracy. Ballbar systems, laser interferometers, and optical coordinate measurement systems are used for specific path and distance tests.

What repeatability values are typical for industrial robots?

Precision SCARA and small six-axis robots achieve ±0.01 mm to ±0.03 mm; standard six-axis industrial robots typically specify ±0.03 mm to ±0.08 mm; collaborative and large-payload robots generally range from ±0.05 mm to ±0.2 mm.

Can repeatability testing be performed on-site?

Yes. Portable laser trackers allow testing at the end user's facility, provided that environmental conditions such as temperature stability, airflow, and floor vibration are controlled and documented. On-site testing by accredited third-party laboratories is widely used for acceptance and periodic verification.

What factors most influence test results?

Key factors include payload and its center of gravity, programmed speed, approach direction, robot warm-up state, ambient temperature, tool mounting stiffness, and the measurement point location within the working envelope. All of these must be fixed and reported for results to be reproducible and comparable.