Spanish efforts to find an electronic alternative to the tried and tested expertise of a human sommelier have now resulted in a system that can tell the difference between varieties of sparkling wine. The new development combines advanced mathematical processing tools with chemical measurement systems and an artificial neural network to create an electronic tongue currently capable of identifying the characteristics of just three cava wines, but with the potential to learn all types available on the market.
Developed by a research team from the Universitat Autònoma de Barcelona's Sensors and Biosensors group, led by Professor Manel del Valle, the Electronic Tongue system uses an array of five graphite-epoxy voltammetric sensors to mimic the sensitive taste system of the sommelier, and detect the amount of sugar added during production. After being cycled in distilled water to get stable readings, the sensors were introduced to 10 ml samples of the 21 varieties of Spanish cava wine (as well as two Champagne samples) and the volumes of information gathered then sent to a preprocessing stage that reduces the influx of complex data into manageable chunks.
UPGRADE TO NEW ATLAS PLUS
More than 1,500 New Atlas Plus subscribers directly support our journalism, and get access to our premium ad-free site and email newsletter. Join them for just US$19 a year.UPGRADE
These chunks are processed by an artificial neural network - now common in numerous in electronic nose and tongue applications - consisting of 235 input neurons which interprets the data.
The system is said to be capable of accurately distinguishing between three varieties of cava wines based on how much sugar was added during the second fermentation stage of the production process - where the bubbles are produced and the sweetness determined. The researchers say that with some training, the system could go on to correctly identify the remaining types, which are classified as follows:
- Brut Nature - less than 3 grams per liter (g/L), no sugar added
- Extra Brut - less than 6 g/L
- Brut - less than 12 g/L
- Extra Dry - between 12 and 17 g/L
- Dry - between 17 and 35 g/L
- Medium-Dry - between 33 and 50 g/L
- Sweet - greater than 50 g/L
In addition to accurately classifying the different sparkling wine varieties, the team says that such a system might also be used to detect defects during the wine-making process.
While it's unlikely that any sommelier worth his salt will be seriously threatened by such a development, it does suggest a possible future where the house wine at your favorite restaurant comes with the approval and recommendation of a cold, calculating, electronic tasting machine.
The paper, entitled Voltammetric Electronic Tongue in the Analysis of Cava Wines by Xavier Cetó, Juan Manuel Gutiérrez, Laura Moreno-Barón, Salvador Alegret, Manel del Valle has been published in the journal Electroanalysis.View gallery - 2 images