Does a robot always hit its repeatability spec?

It holds, but the margin disappears. A study of a modular arm rated at plus or minus 0.1 mm found the worst axis spread 0.0879 mm at 6 kg and about 0.2 mm at 16 kg. Since a plus-or-minus 0.1 mm rating allows 0.2 mm of total spread, the heavy load put the arm at the edge of its band rather than outside it.

Why does load affect repeatability?

By ordinary mechanics: a heavier payload deflects the arm and joints more and works the drivetrain harder, so the position the robot returns to varies more. The paper measures the result rather than attributing a cause, and what it shows is that accuracy erodes as the load approaches the rated maximum.

Do these numbers apply to any robot?

No. They are one modular arm in one study. The direction, worse repeatability under heavier load, transfers; the exact 0.2 mm does not.

How should I specify a robot for accuracy?

Treat the datasheet repeatability as a light-load best case, plan for degradation near the rated payload, and leave margin rather than assuming the headline number under full load.

Robot performance

Robot repeatability under load: the spec versus the measurement

A robot arm rated for plus or minus 0.1 mm repeatability measured about 0.2 mm of spread under a 16 kg load, filling its entire rated band with nothing to spare.

The figure comes from a study that measured positioning accuracy across loads of 6, 10, and 16 kg. At 6 kg the worst axis spread 0.0879 mm, comfortably inside spec. At 16 kg the range on two axes reached roughly 0.2 mm, which is the full width a plus-or-minus 0.1 mm rating allows.

This page traces the measurements to the study and makes a narrow, useful point: a repeatability spec is a single condition, and the number you actually get depends on the payload.

Data covers Measured positioning study of a modular industrial arm under varying loads (2025). Published 2026-07-14. Last reviewed 2026-08-16. Last updated 2026-08-16. Edited by Mike Ramsey / Reliable Media.

The figures and where they come from

Each figure is rated for how safely you can cite it today. Ratings judge current usability, not whether a number was ever correct.

Figure	What it is	Source	Citation Confidence	Notes
plus or minus 0.1 mm	Rated repeatability	[A]	High	The arm's datasheet repeatability specification. A single number, quoted as if it held across all conditions.
0.0879 mm	Measured range at 6 kg (worst axis)	[A]	Medium	At a 6 kg load the worst axis spread 0.0879 mm, well inside the 0.2 mm of total travel a plus-or-minus 0.1 mm rating allows. Light loads meet the datasheet with room to spare.
about 0.2 mm	Measured range at 16 kg	[A]	Medium	At 16 kg the range on the Z and Y axes reached about 0.2 mm. A plus-or-minus 0.1 mm rating permits exactly 0.2 mm of total spread, so the arm arrived at the edge of its band with no margin left.
6, 10, and 16 kg	Loads tested	[A]	High	The study measured across three payloads, which is what let it show accuracy changing with load rather than reporting one condition.

Why the numbers disagree

A datasheet repeatability figure is measured under specific, favorable conditions, usually a light or nominal load. The study shows why that matters: the same arm that spread 0.0879 mm at 6 kg spread about 0.2 mm at 16 kg, which is the whole width a plus-or-minus 0.1 mm rating permits. The spec is not wrong. It just describes one load, and it is a limit rather than a promise of margin.

Repeatability and payload are linked by ordinary mechanics: a heavier load deflects the arm and its joints more and works the drivetrain harder. The study measures the effect rather than explaining it, so treat the mechanism as standard engineering knowledge and the numbers as the sourced part. Either way, what a robot delivers in a real cell depends on how close the application runs to the rated payload, not just on the headline spec.

This is one arm in one study, a modular industrial robot, so the exact figures are specific to it. The general point, that measured accuracy degrades toward the edge of spec as load approaches the maximum, is the transferable lesson, not the precise 0.2 mm.

How to cite these figures

Read a repeatability spec as a best-case, light-load figure, and expect degradation as the payload approaches the rated maximum.

If accuracy matters in your application, cite the measured behavior: this arm spread 0.0879 mm at 6 kg and about 0.2 mm at 16 kg, using up its rated band. Plan for margin rather than assuming the datasheet number leaves any at full payload.

Treat the exact 0.2 mm as specific to the tested arm. The transferable claim is directional: heavier load, worse repeatability.

Where people go wrong

Assuming the datasheet repeatability leaves margin at full payload. It is typically a light-load figure, and here the measured spread more than doubled between 6 kg and 16 kg, arriving exactly at the edge of the rated band.

Generalizing the exact numbers to any robot. They are one modular arm in one study; the direction transfers, the values do not.

Ignoring payload when specifying for accuracy. Running near the rated maximum is where repeatability erodes.

How we checked

The figures come from a peer-reviewed positioning-accuracy study that measured a modular industrial arm under three loads. We accessed it through PubMed Central and confirmed the plus-or-minus-0.1 mm spec, the 0.0879 mm range at 6 kg, the roughly 0.2 mm at 16 kg, and the tested load set in its text. One caution about the comparison: the datasheet quotes repeatability as plus or minus 0.1 mm, while the study reports a per-axis range, the distance between the highest and lowest measured value. We read them against each other by treating the rating as the 0.2 mm band it describes, which is the like-for-like comparison. The two are not defined identically, so treat the margin as indicative rather than as a conformance test.

We frame it as a single measured study whose value is the load comparison: measuring at 6, 10, and 16 kg is what reveals accuracy changing with payload, something a single-condition test would miss.

The precise numbers belong to the tested arm. We state the transferable finding, repeatability degrades as load approaches the maximum, and hold the exact values to their source rather than presenting them as universal.

Full source list

Primary sources, with live links. Every figure above traces to one of these.

[A]PMC (PubMed Central)2025
"Study of Positioning Accuracy Parameters in Selected Configurations of a Modular Industrial Robot, Part 1", PMC (PubMed Central)
https://pmc.ncbi.nlm.nih.gov/articles/PMC11722867/

Common questions

Does a robot always hit its repeatability spec?: It holds, but the margin disappears. A study of a modular arm rated at plus or minus 0.1 mm found the worst axis spread 0.0879 mm at 6 kg and about 0.2 mm at 16 kg. Since a plus-or-minus 0.1 mm rating allows 0.2 mm of total spread, the heavy load put the arm at the edge of its band rather than outside it.
Why does load affect repeatability?: By ordinary mechanics: a heavier payload deflects the arm and joints more and works the drivetrain harder, so the position the robot returns to varies more. The paper measures the result rather than attributing a cause, and what it shows is that accuracy erodes as the load approaches the rated maximum.
Do these numbers apply to any robot?: No. They are one modular arm in one study. The direction, worse repeatability under heavier load, transfers; the exact 0.2 mm does not.
How should I specify a robot for accuracy?: Treat the datasheet repeatability as a light-load best case, plan for degradation near the rated payload, and leave margin rather than assuming the headline number under full load.

More data, traced to source

Robot reliability numbers: the vendor claims and the one independent study
Manufacturers advertise robot uptime in the high nineties and mean time between failures in the tens of thousands of hours. The one independent study of more than 400 factories found a robot cell is reliable 88 percent of the time, with 87 minutes between failures.
How much energy does an industrial robot use? Mostly overhead
Two measured studies of industrial arms found that overhead, electronics and brakes and holding position, dominates power draw, and that less than 2.5% of the energy consumed becomes useful mechanical work. The numbers vendors do not publish.
A cobot's measured productivity gain: 10%, not a multiple
Collaborative-robot marketing implies large throughput gains. A time-studied assembly cell measured the actual gain at 10%. Here is the study, and why a measured number beats a marketed one.
A robot's repeatability is not its accuracy: the measured gap
A robot's repeatability spec is a sub-millimeter number; its actual accuracy out of the box is far worse. In one measured study a high-precision arm was off by 1.7730 mm before calibration, and calibration cut that to 0.1041 mm.
How well do robots actually pick from a bin? The measured rates
Robots do not pick perfectly. On real hardware, a state-of-the-art suction system succeeded on 82.3% of picks, and just 73.3% on hard objects. A separate system reached 600 picks an hour at 96 to 99% on dense metal parts.