Creating a Metrology Taxonomy

by Michael Schwartz & David Zajac

Metrology is the science of measurement. For thousands of years we have been creating newer and better ways of measuring things, but in all that time, metrology still lacks a detailed and standardized taxonomy. Just like the biological and medical sciences, I believe this is something that metrology science would benefit greatly from, if it were created and adopted. This is especially true now, as we integrate information science with metrology, we have determined that the standardization of taxonomy for metrology is a basic requirement for the efficient exchange of measurement data.

Taxonomy is at the top of our list for our company because it is needed for Metrology.NET® to have the ability to check measurement uncertainty calculations against a lab’s Scope of Accreditation (SOA). We are creating a system where every measurement in a calibration can be verified against a lab’s accredited capabilities. This requires the calculated uncertainty be checked against the lab’s accredited capabilities, then for the test report, choosing between the larger of the calculated or accredited value, and maintaining all measurement data to prove the lab’s best uncertainties during the next audit!

To do this, we must first define a way to link every measurement to the correct section of the lab’s SOA. The system must be able to search and select the correct information from the SOA each and every time! There can be no ambiguity in the interface between the information contained in the SOA and our system!

It is important to note, Cal Lab Solutions has been working on this effort for almost three years now. Most recently, in 2016 we presented a paper on the topic, “Creating a Standardized Schema for Representing ISO/IEC 17025 Scope of Accreditations in XML Data,” at NCSLI Workshop & Symposium, St. Paul, MN. Now we are working in synergy, as part of the Metrology Information Infrastructure (MII). Since the MII meeting in St. Paul, several companies (like Boeing and Qualer) have taken leadership roles in further defining a machine readable SOA. [For further reading about the MII meeting activity >> “Toward a Measurement Information Infrastructure: Setting Sail.”]

What we discovered fairly early in developing a prototype search tool—code named Beagle—was we needed a quick and easy way to index the measurement category before we could define the values required to search the SOA. For example, if you are searching for a lab’s Watts measurement capabilities, you can’t just search on Watts values. Metrology information systems are complicated by factors such as Watts can be the product of Amps and Volts. If we are looking for AC watts, then we need to include frequency; if we are looking at Watts that incorporate Power Factor, we also have to include Phase Angle. Simply searching for Watts without taking into consideration these potential complicating factors can be insufficient and can return unwanted values.

We have created a solution that relies on a robust and standard metrology taxonomy to create a standardizable, hierarchical information backbone for organizing all of metrology’s sub categories and subtle variations. This reduces the search complexity by orders of magnitude! Thus, creating a standard for definition metrology based taxonomy is imparative for both Metrology.NET and the MII project.

Our proposed standard will define a syntax for a naming convention with increasing specificity. For example, Measure.Watts.AC or Source.Volts.DC; each dot(.) divides the branch into a more specific subcategory. Every leaf of the taxonomy tree would then contain a parameter list used for sorting and filtering. So for these examples, Measure.Watts.AC would define parameters for Amps, Volts, Frequency, Power Factor, as well as Watts as search inputs. And Source.Volts.DC would have Volts and maybe Input Impedance. This Metrology Taxonomy definition would be used to categorize the specific hardware and technique implementation. A couple of examples would be Source.Volts.DC(5720) and Source.Volts.DC(5520.Characterized.3458). By indexing both of these specific implementations to a specific leaf on the metrology taxonomy tree, they become equally searchable yet categorically distinguishable.

We will be presenting our version of the Metrology Taxonomy model at the MSC Training Symposium in Anaheim, CA, in April and hope to present a paper relating to the topic at NCSLI 2017 in National Harbor, MD.