Have you ever wondered whether you NIR results are trustworthy, or how well they compare to reference analysis? Based on many years of working with NIR in the international feed, food and laboratory industry, Key Account Manager Christian Tollebäck reflects upon the quality of NIR results and the tools to accredit them.
In 2010 when I first became aware of the ISO 12099 for NIR, Animal feeding stuffs, cereals and milled cereal products, Guidelines for the application of near infrared spectrometry, I strongly believed this would be the final acceptance of NIR as a standard method for quality control of feedstuffs.
Of course NIR has been used in the feed industry as well as other industries for many years and many companies have been using it as their main analytical method, using wet chemistry only as back up and for validating their calibrations. Therefore, it is interesting to follow how more and more companies are on their way to accredit their NIR methods under ISO 17025 and some have already succeeded.
Talking to people who either had a bad first experience with NIR, or who have little to no experience at all, a very common question I hear is: How can I trust my NIR if the result is different compared to my Kjeldahl analysis? My first response to this question is: How can you trust your reference method?
Better performance than the reference method?
Yes of course NIR is a secondary method, but there is no doubt that the reason we can measure protein is that the energy absorption in the NIR region gives us real information about the protein content. Could NIR be even better than the wet chemistry method? Let’s take a theoretical example; assume that we are able to control all the variations in the sample being measured. Then we can assume that the error in the method comes mainly from the reference method and to some extent from the noise of the instrument. Building calibrations is a costly task as you need fairly large databases with good reference data. Many NIR users duplicate their reference method in order to have more accurate results as they know how important it is to have quality data in their NIR calibrations. What would happen if we had the luxury of using 100 reference results and taking the average of those 100 results for each sample in our calibration? The average of these 100 results would then represent the true value. Would this calibration show better performance than the reference method itself?
Obviously, this is only a theoretical discussion, but it is clear that there are now good tools to accredit your NIR results. Today’s modern instruments combine a high level of sensitivity, transferability, quality control, networking software and other technical improvements that help to build trust in the NIR results, which is why we can now use the Term “plug and play” with confidence.
Want to know more about the potential and application of near infrared technology in feed production? Join our fast growing network of experts and professionals from both NIR and feed sectors now.