New technologies are steadily increasing our grain analysis capability. The challenge for FGIS, and by extension, for all participants in the grain market, is to incorporate new technologies or improved operating procedures without creating discontinuities or disruptions in markets. Analytical differences can create instant economic stress if products are revalued or misvalued among market participants.
FGIS goes to great effort to standardize test results across inspection points, not only on average performance, but on individual sample to sample results. Especially for calibrated, often electronic, tests, the inherent differences in response of different measuring systems, like moisture meters, NIR analyzers, or test weight systems creates statistical variability across testing locations, variability that potentially increases the natural variability of a single instrument system, This is the reason that FGIS limits the number of makes and models of instrumentation for a given test – to control any consistent differences across technologies even though on average each technology is accurate related to the fundamental reference for the test.
Exclusive use of a single make and model of instrument naturally creates protection for that instrument; examples are one NIR unit for composition testing, two units of the same test cell design for moisture, or one sieving device for particle size factors. In 2014, the GIAC asked ISU to determine for one test, NIR composition, how large the variability might be for multiple instruments compared to the one instrument that FGIS was (and still is) using. In a report in 2016, the conclusion was that when multiple NIR instrument models were calibrated to the FGIS reference lab on the same calibration sample sets, the variability (standard deviation across instruments of wheat protein measurements) was only 0.01 percentage points greater for all models (3) pooled than for the Official instrument copies alone.
Recommendation: Develop a proposed protocol for measuring equivalence that can be applied to potential alternative, new technologies under consideration for Official use. The general criterion would be that the use of additional technologies for a given test would not create unacceptable systemic variability beyond what is present with the existing technology.
Initial Author:
Dr. Charles R. Hurburgh
Professor, Agricultural and Biosystems Engineering
Professor in Charge, Iowa Grain Quality Initiative
3167 NSRIC, 1029 N. University Blvd.
Iowa State University
Ames, Iowa 50011
515-294-8629
tatry@iastate.edu