Every physical measurement in research and industry involves some degree of uncertainty. Even minor fluctuations in ambient conditions can influence the measurement and lead to deviations that make the measurement result unreliable. Knowledge of the uncertainty pertaining to the results is essential in order to obtain meaningful measurement data.
![[object Object]](https://kistler.cdn.celum.cloud/SAPCommerce_CMSGalleryStandard_720x480/1416-graph-in-curves.webp "Graph in curves")
The measurement uncertainty quantifies the insufficiency of knowledge about the true value of the measurand. It is a parameter [u=uncertainty] assigned to the measurement result as a percentage, indicating the spread of values that can be reasonably attributed to the measurand on the basis of the available information. In other words, it is a range within which the true value of the measurand is located, with a defined degree of probability.
There are various ISO-standardized methods for determining the measurement uncertainty. Probability distributions provide the basis for the analytical/computational method as per ISO/IEC Guide 98-3. It is based on quantitative information, and also on non-statistical parameters such as empirical values or data from a calibration certificate. Calculation of the measurement uncertainty depends on the measuring task and the measuring principle as well as the measurement process, and it takes account of all relevant input variables that could influence the measurement. It is therefore the result of a combination of all influencing factors of the respective input variables. The more elements a measuring chain contains, the more complex it becomes to determine the measurement uncertainty.
Another method of determining the measurement uncertainty is the collaborative study (or 'round robin') procedure as per ISO 21748. This involves multiple laboratories that investigate identical specimens using the same measurement methods. This is followed by a comparison of the measurements within one laboratory and the measurements between the respective laboratories, resulting in a standard uncertainty for the measurand under investigation.
Lack of information about the uncertainty of the measurement result can cause misinterpretations. In the industrial context, this could result in incorrect decisions leading to unnecessary costs, image loss and even claims for compensation or criminal prosecution.
Reliable determination of the measurement uncertainty allows narrower tolerances to be defined so the efficiency of the process can be increased.
