Wine is an incredibly complex substance whose production is an equally complex process influenced not only by nature, but also by viticultural and winemaking practices. It is estimated that wine is composed of up to one thousand molecules which, in concert, give each wine its idiosyncratic odours and flavours. The precise combination of molecules that is ultimately present in a bottle of wine is affected by everything from the soil composition, rainfall, and sunshine hours of a particular region, to the rootstocks, maturation vessels, and bottle closures chosen by the winemaker. Some molecules are even synthesised after bottling, either as a result of winemaking decisions before bottling, or as a result of storage and transport conditions.
In wine tasting, a wine’s characteristics are usually grouped into categories that make it easier for a human palate to draw conclusions about its provenance, quality, production, and appeal. Thus, when we taste a wine analytically, we try to evaluate its aromas, flavours, sweetness, acidity, tannins, alcohol, and body in order to draw conclusions about where and how the wine was made, how old it is, what grape(s) it was made from, whether it is faulty and ultimately, whether it is pleasant. What we are technically and intuitively doing with our noses and palates is detecting the presence and proportion of alcohols, acids, phenols, monosaccharides, esters, thiols, aldehydes, pyrazenes, and other molecules in our glass of wine. Some people become so good at this exercise that they can spend a few minutes with a glass of wine and correctly identify the producer, vintage and price of a wine. This can seem like magic to the inexperienced taster, but is really nothing more than rigorous detective work: the methodical deduction of attributes from olfactory and gustatory clues.
One of VeriVin’s ambitions is to go beyond fault testing and eventually attempt to replicate and extend the capabilities of a human taster by partially – or even fully – characterising the molecular makeup of a bottle of wine. Some implications might be practical. Imagine a comparative purchasing tool that could give a broad measure of the acidity and tannic grip of two different bottles of wine on a scale of one to ten. Some implications might be downright futuristic. Imagine being able to non-invasively determine the precise and complete molecular makeup of a bottle of wine: you could then use that data along with data from other bottles (and a machine learning algorithm) to predict its Wine Spectator score.
Back to Our Technology & Applications.