Continuing the quest for the magic 'black box'
May 01, 1999
by Teresa Acklin
The single kernel characterization system is finding acceptance in the wheat milling industry as a valuable tool for wheat selection, blending, and mill optimization
By Jim Psotka
The search for the elusive “black box” began many years ago. As new technologies such as near-infrared spectroscopy were developed in the latter part of this century, there arose the desire to develop a system that could analyze a sample of grain and provide physical and chemical information about the grain quality and how the grain could be used.
The goal of such technologies became and continues to be something of a moving target. This is partly because of the cost of the technology, the continuing evolution of technologies and the development of new instrumental techniques, such as image analysis and faster computers. One such development in the pursuit of the “black box” is the single kernel characterization system (SKCS).
Since its introduction in 1993, the SKCS is finally finding acceptance in the wheat milling industry as a valuable tool for wheat selection, blending and mill optimization. Having once been applauded as the most significant development for the milling industry of the decade, it has taken considerable time for the industry to accept its utility.
The apparent delay in acceptance of the SKCS is attributable to several factors: perceived limitations to the utility of the instrument, a lack of understanding of the information provided by the instrument and restrictions of the instrument to small cereal grains. While these appear to be negatives of the technique, the expanding user base is indicative of the gradual realization of its value.
In its original design, the SKCS was intended to be used by the Federal Grain Inspection Service (FGIS) as an objective means of classifying North American wheat as hard or soft and for identification of mixed classes. The need for objective classification arose as new wheat varieties were developed that were difficult to classify visually.
Field testing the instrument with FGIS classification algorithms demonstrated that the hardware was sufficiently rugged to sustain heavy use, especially during the harvest season. However, the classification algorithms used exhibited some difficulty in distinguishing hard and soft classes for some wheat varieties, resulting in incorrect classification or misclassification as “mixed” wheat.
While this difficulty can be corrected in the instrument, implementation of the SKCS system for classification has been slow. Thus, many potential users outside of official inspection agencies of the instrument have been slow to adopt the technology.
To further complicate matters, the use of the Hardness Index scale, required for correct classification, often does not rank wheat varieties in the same manner as do other hardness methods. That is, other hardness methods, such as NIR hardness, softness equivalence or Particle Size Index (PSI) hardness, may not agree with the SKCS on wheat varieties that are apparently harder or softer.
The SKCS provides results for 300 kernels measured. The average, or mean, kernel weight, kernel diameter, kernel Hardness Index and kernel moisture are displayed. The variability (standard deviation) of each of the four parameters also is displayed.
The variability is an indication of the uniformity, or lack of uniformity, within the sample tested. The uniformity of kernel weight and diameter are related to first break flour release and milling yield. Hardness Index average and uniformity are related to conditioning requirements and first break release. Moisture average and uniformity are important to wheat tempering and conditioning.
While these are probably simplistic assessments of the role of these variables, they serve to make a point. Historically, information about within-sample variation required time-consuming physical measurements, which generally rendered such measurements impractical for the mill. The use of the SKCS has enabled researchers and users to demonstrate how to use these measurements in optimizing grain cleaning and separation, wheat blending and wheat conditioning.
One of the bigger disappointments to researchers is the limitation of the SKCS to small grains. Researchers using the SKCS have reported on work with wheat, barley, barley malt, sorghum, oats and rice. In these cases, the SKCS has provided useful information that was previously difficult to obtain, such as single kernel weight, diameter and Hardness Index.
Moreover, information about the uniformity of the grain is usually very useful. It is, however, frustrating to know that the system cannot be used for whole soybeans or kernels of corn or other large grains or for very small grains such as rapeseed.
A “hardness” measurement for corn kernels would be very helpful, for example, to assess internal damage in corn due to handling or drying practices. While the SKCS cannot handle these larger kernels at the present time, it is certain that instrument manufacturers will resolve this issue.
Still, there is a bright future for SKCS technology. Fortunately for the grain industry, the SKCS has been demonstrated to have application well beyond the narrow focus of Hardness Index or classification. In addition, research is continuing to find new ways to utilize the SKCS by creating predictive models.
The system is permitting millers to look at grain on a single kernel basis and to more efficiently look at the impact of kernel variability on a number of quality factors in grain handling and milling. The information is forcing some new thinking in order to understand the significance of SKCS and how to best convey this message to the industry.
GRAIN HANDLING APPLICATIONS.
The instrument provides a means of detecting hard or soft wheat blends or blends of wheat with different moisture levels. The SKCS has also been used to screen for and reject potentially poor quality wheat shipments, based on SKCS prediction models. This prevented receipt of these shipments and subsequent need for blending into bread wheat mixes or disposition of grain unusable for the production of bread flour. Researchers using the SKCS have shown how information from the instrument can be modeled to provide information related to wheat milling performance. An estimation of flour yield can be made that show differences in milling properties that will directly affect yield.
This milling index is applicable to both hard and soft wheat varieties. This information is useful not only to the millers, but also to wheat purchasers since wheat can be selected based on SKCS milling index as well as other parameters.
Similar measurements have been developed for durum wheat for estimating semolina yield. Further, this type of data is of value to wheat breeders in selection varieties for development.
Studies have shown the value of SKCS information in mill optimization using sizing or blending of wheat. The instrument also has been shown to be useful in evaluation of tempering and conditioning parameters.
The way to overcome obstacles to implementation of this technology is to educate the industry about what the instrument can do now and what direction researchers can take to facilitate acceptance of this technology. Without sounding like a late-night television “infomercial,” part of the challenge for researchers is to “spread the word” about the SKCS by providing concrete examples of the utility of this technology in the grain industry and discuss what direction the research needs to take to expand the industry's use of the technology.
This interaction between the grain industry, researchers and instrument manufacturers is critical to the evolution of technologies to finally deliver the magical “black box,” or at least the next step toward it.
So, what about the future? Many of the questions raised in original discussions of the need for the “black box” remain either unanswered or inadequately answered. Research has come a long way with analytical technologies, and the goals have been redefined again.
Researchers want more and better information with which to make decisions. One of the answers to this pursuit may be the combination of technologies in a single instrument. The SKCS is already a hybrid, combining sensors for weight, crushing force and conductance to provide information about kernel weight, diameter, Hardness Index and moisture. Current development of a near-infrared (NIR) module for the instrument will greatly enhance its capability for evaluating other grain characteristics, such as color and vitreousness, as well as less obvious assessments of quality, such as internal insect infestation in kernels and detection of bunted kernels.
Beyond these, the combination of physical data with chemical, or spectral, data and the use of more sophisticated algorithms for calibrating the instrument will greatly enhance the capability of the system. It suffices to say that the industry is ready for the next lap in the pursuit of the “black box.”