What type of beer does your robot prefer?

Spanish researchers have managed to distinguish between different varieties of beer using an electronic tongue. The discovery, published in the journal 'Food Chemistry', is accurate in almost 82 percent of cases. Credit: Manel del Valle

Well, actually, how could a robot like one kind of beer better if it cannot distinguish one from the other? 

Right, but wrong! Apparently, Spanish researchers have managed to develop a "sense of taste" that can serve well a robot in distinguishing with an accuracy of over 80% a beer (brand, type....) which is much much better than what I can ever do!

Clearly one could imagine that with all the chemical analyses tools at our disposal it shouldn't be that difficult to identify a specific type of beer with good accuracy. But chemical analyses is a long process and just sipping wouldn't do. What researchers at the Autonomous University of Barcelona did was to develop electronic tongues consisting of an array of sensors containing 21 ions-selective electrodes. Some electrodes are able to detect ammonium and sodium, others can detect nitrate and chloride and other are responsive to general molecules.

The process leading to the identification of the beer mimics what happens in our brain when we taste something and it involves a lot of software, in our case neurones and in this case intensive processing through microchips.  Our taste buds react to some specific molecules and send the messages to the brain that interprets these signals (along with the one coming form the nose...) raising the perception of a specific taste than is therefore compared to previous experiences of that taste to lead to the identification of the beer.

The system developed works in a very similar way, and as we do, it learns new tastes getting better and better. And, as it happens to us if we want to become expert in wine  or beer tasting, you need to be coached by an expert in the field that tells us what to pay attention to (that sweetness emerging after tasting, the fruity flavour....). 
Researchers did just that. They supervised the tasting providing hints on what to pay attention to and provided connections among different categories: Schwarzbier, lager, double malt, Pilsen, Alsatian and low-alcohol.

Apparently, wine tasting is more complex than beer and more work will be needed. The idea is not to substitute a "sommelier" with a "robot" but to provide robots with a sense of taste that may be useful in performing certain tasks.

Interestingly, information technology is playing a crucial role, both in the learning and in making a sensation emerge out of a variety of signals.

Author - Roberto Saracco

© 2010-2018 EIT Digital IVZW. All rights reserved. Legal notice