30H-2 |
Comparison of sensory and consumer testing with E-nose and E-tongue for fruit juices |
T. T. TAN1, S. Isz2, S. Labreche2, E. Saint-Martin2, R. Bleibaum3, and H. Stone3. (1) Marketing Dept., Alpha M.O.S, 20 Ave. Didier Daurat, Toulouse, 31400, France, (2) France, (3) Marketing Dept., Tragon Corp., 365 Convention Way, Redwood City, CA 94063-1402 The use of sensory panels in manufacturing plants is significantly more difficult than within a research and development environment. The production environment is very hostile to sensory analysis. The noise, odors and pressure make it extremely difficult to get objective and reproducible results. Certainly, there are strong requirements to maintain and improve current sensory QC/QA methods that are carried out by panelists. Alternative or complimentary methods must be fast, accurate and cost effective. Of course, the results must be correlated to the results of sensory analyses. The objective of this research was to determine if results from electronic tongue and nose (sensors and mass spectrometry) have high correlation with those obtained from a qualified sensory panel using the QDA method (Stone and Sidel, 1993) and consumers (acceptance). The results will indicate if these instruments can be used to predict/monitor sensory characteristics, which leads to consumer acceptance. Electronic noses, electronic tongues, expert sensory panels and consumer panels were used to evaluate commercially available apple juice. A series of nine 100% apple juices were tested including three-apple blend, vitamin C fortified, apple pear juice, and an apple cider. All but one product base was Tree Top brand apple juice, one of which was altered to analyze increases in sweetness (sucrose) and sourness (citric acid). Relationships between the QDA means (34 attributes) and instrumental measurement means (25 sensors) were identified using principal component analysis (PCA). Based on the PCA without Sour 1, a wheel plot was created to understand the relationships more easily. Good correlation were found between consumer, QDA and electronic sensor measurements. The results showed that electronic tongues and electronic noses may be used to track "consumer-defined quality" of apple juice by using results from correlation's and predictive equations between QDA and consumer data.
Session 30H, Quality Assurance
|