Analytica Chimica Acta, 284 (1993) zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA l-11 Elsevier Science. Publishers B.V., Amsterdam Quality classification of grain using a sensor array and pattern recognition J.R. Stetter, M.W. Findlay, Jr., KM. Schroeder, C. Yue and W.R. Penrose Transducer Research, Inc., 999 Chicago Avenue, Napervilk, IL 60540 (USA) (Received 15th September 1992; revised manuscript received 13th July 1993) Ahstmet Measurements using arrays of electrochemical gas sensors, combined with pattern recognition methods, were used to classify wheat samples by quality grade. The classifications corresponded closely to those made by trained grain inspectors. Volatile compounds evolved from warmed samples of grain were passed over a heated noble metal catalyst and then into a series of electrochemical sensors. Signals from four sensors were recorded for four different catalyst temperatures in order to generate 16 signals for each grain odor sample. The 16 sensor signals were treated as a 16-dimensional vector or pattern of responses that was characteristic of the odor sample. The patterns for different grain odor samples were compared using both nearest-neighbor analysis and a commercial neural network simulation (NNS) program. These methods classified the samples correctly by grade with an accuracy of 68% and 65%, respectively. After compensation for instrument parameters, the NNS score improved to 83%; the nearest- neighbor analysis could not be similarly compensated. The robustness of the two algorithms was compared by adding simulated random and systematic errors to the sensor response patterns. The original data were used as the training set, and the patterns with errors added were used as the test set. In these cases, the NNS consistently outperformed the nearest-neighbor method at classification of the grain odor samples. Kcywordr: Pattern recognition; Sensors; Grain; Quality classification All grain exported from the United States must by law be inspected by agents of the Federal Grain Inspection Service (FGIS), US Department of Agriculture, while being loaded into the ships, holds. Additionally, there are inspectors who are certified by the FGIS who may be hired to in- spect grain during domestic transfers of custody. A subjective judgment of odor is used by the FGIS as a primary criterion of the fitness of grain for human consumption. An experienced inspec- tor inhales from a sample of the grain and evalu- ates the grain as Good, Sour, or Musty. Inter- nally, additional designations, such as COFO (commercially objectionable foreign odor) are also Corrarpondence to: M. Fmdlay, Jr., Transducer Research, Inc., 999 Chicago Avenue, Naperville, IL 60540 (USA). used. Should a sample be classified as bad due to odor or any of the other parameters being moni- tored, loading of the ship is halted, and the grain already on the ship must be unloaded. Because these evaluations frequently involve large sums of money, there is an appeal process which delays and confuses the task of grain grad- ing. A sample of the grain in question is sent to the FGIS Board of Appeals and Review, where a panel of inspectors grades the sample. The final classification is the average of the individual judg- ments of the inspectors. Subjective judgments are difficult to defend in an adversarial environment, and the USDA has searched for more objective means of evaluating the quality of a sample of grain. It is recognized that this is a philosophical contradiction, since 0003-2670/93/$06.00 Q 1993 - Elsevier Science Publishers B.V. All rights reserved