Research And Application Of Smartongue And Smartnose In Brand Identity And Quality Stability Of Coconut Milk

In recent years, as people increasingly advocate a green and healthy lifestyle, vegetable protein drinks and tea beverages get a widely love by consumers. However, homogeneity, brand clutter, price war and the lack of innovation have become increasingly prominent. It requests that the enterprises not only need a right broad promotion, but also a quality control of the products. From the field to the table, each link in the chain needs strict control. In the beverage industry, quality testing instruments just detect the individual index of the product, neglect the interaction between factors, and lack a comprehensive evaluation. The sensory evaluation, vulnerable, time-consuming, limited, cannot meet the requirements of modern industry. A combined system for coconut milk quality monitoring is proposed in this paper. It consists of a potentiometric electronic tongue based on multi-frequency large amplitude pulse voltammetry (MLAPV) and a humid electronic nose, to establish a set of intelligent sensory evaluation and comprehensive quality evaluation of coconut foods to ensure food safety. The conclusions go as follows.1Compared with the two dimension results presentation, a three dimensional structure can cover the shortage of fixed projection viewing perspective adopted in the previous Smartongue data processing. Then convert the three dimensional coordinates to two, as human vision is a two-dimensional plane. In addition, the discrimination power is used to evaluate the identification rate between individuals.2Smartongue has the capacity to identify the brands and the formulas of coconut milk, with the DI value of90.1%. It turned out that the coconut milk of the same brand was distributed in a certain area. One can identify the product authenticity according to the database, with a total recognition accuracy rate of94.44%. The discrimination between different batches of coconut milk implies that the differences occur in the overall taste characteristics.3Smartnose has the capacity to identify the brands and the formulas of coconut milk, with the DI value of89.4%. It turned out that the coconut milk of the same brand was distributed in a certain area. One can identify the product authenticity according to the database, with a total recognition accuracy rate of88.89%. The discrimination between different batches of coconut milk implies that the differences occur in the overall odor characteristics.4The combination system of Smartongue and Smartnose, through data fusion has more to do with the results of Smartongue, but less to do with that of Smartnose. Due to the difference of sensory stimulation between the smell and taste in coconut milk, Smartongue and Smartnose show their characteristics, respectively.As can be seen from the results of this study, Smartongue and Smartnose have a good ability to recognize coconut milk drinks. The application of Smartongue and Smartnose in coconut milk suggests that these intelligent technologies might be extended to other vegetable protein beverage’s quality control. With the advantages of rapid, non-destructive, real-time monitoring of the production process, Smartongue and Smartnose can cut losses and waste of resources, for promoting economic and social benefits of enterprise.