Repeatability vs. Accuracy in Inline 3D Measurement

Sensors used for inline inspection must be both repeatable and accurate to be effective for quality control and process monitoring. Without suitable accuracy, measurement data cannot be trusted for decision-making.

Accuracy and repeatability can be illustrated by considering shots at a target, with the center of the target representing the true value of the desired measurement.

Accuracy versus Repeatability


Standard sensor industry practice defines accuracy (also called linearity) as the largest difference from the “true value” and the measured value, over the complete measurement range. Accuracy can be stated in engineering units, or as a percentage of the full-scale measurement range.

Verifying accuracy of a sensor is a complex process. It is usually determined from test procedures, which involve moving an object or target in small steps through the entire measuring range, with the target position measured by an external means (such as a laser interferometer or precision glass scale). Differences in the target position as measured by the sensor are compared to the measurement positions determined by the external device to determine the sensor accuracy.


Repeatability is defined as the variation in output data over time from a stationary or fixed target.

Verifying sensor repeatability is easy. Simply place a stationary target in front of the sensor, and observe the variation in sensor output over time, through the range of environments (such as ambient temperature) the sensor will encounter.

Repeatability is an important value in sensor specifications since a sensor must be repeatable in order to be determined as accurate. In fact, repeatability values must be smaller than accuracy values by definition.

Accuracy and Repeatability – Not the Same Thing

It is important to remember that accuracy and repeatability are not the same. A sensor may show extremely good repeatability when set up on a stationary target, but could be very non-linear and not accurate. Repeatability only shows that the sensor gets the same reading from a stationary surface, time after time – it does not prove that the measured value has any relationship to the actual value.

From a user point of view, repeatability, unfortunately, sometimes receives more emphasis than it deserves, doubtless because it is verified by a quick and simple test that does not require any special equipment or procedures. Accuracy is more difficult for a user to verify, but is critical to ensure measurement data is valid. A common practice is to use a sensor that has accuracy 10 times smaller than the targets specified tolerance.

A simple way to ensure accuracy is to implement the best of today’s smart 3D sensors, which have accuracy verified by the manufacturer with precision calibration equipment, and are designed to maintain accuracy over the range of environments encountered in industrial applications.