Could an electronic nose developed by Nestlé be the ultimate coffee snob of the future? Company researchers say the ability of their machine to differentiate among flowery, winey, bitter, cocoa and other espresso attributes could help with quality control.
-- Bryn Nelson"; Tech_Frontiers_Espresso_080310[i++] = new Array("","Human vs. machine","","http://msnbcmedia.msn.com/j/msnbc/Components/Photo_StoryLevel/080227/Espresso-1.hmedium.jpg","","", "", "", "", "", "", "Courtesy of the Alliance for Coffee Excellence", "", "273", "333", "#000000", "", "", "", ""); Tech_Frontiers_Espresso_080310[i-1].body = "Human vs. machine
The machine assembled by Nestlé scientists isn’t intended to replace human judges, like this one carefully sniffing a coffee sample in the 2005 Nicaragua Cup of Excellence competition. But in an analytical test of the machine’s abilities, it successfully predicted the range of aromas and intensities noted by a panel of experts for 11 different espressos, with few mismatches."; Tech_Frontiers_Espresso_080310[i++] = new Array("","Espresso attributes","","http://msnbcmedia.msn.com/j/msnbc/Components/Photo_StoryLevel/080227/Espresso-2.hmedium.jpg","","", "", "", "", "", "", "Courtesy of Analytical Chemistry", "", "273", "268", "#000000", "", "", "", ""); Tech_Frontiers_Espresso_080310[i-1].body = "Espresso attributes
Tasked with judging different espressos on attributes ranging from flowery and winey to cereal and woody, the predictions made by Nestlé’s electronic nose closely matched the collective opinion of 10 human panelists for most espressos. Results for six of the 11 samples are shown above. Each colored shape indicates the machine’s predicted sensory profile (the more intense the attribute, the more the color extends from the middle), whereas the red outlines show the panel’s conclusions. Some samples, such as Espresso #1, show especially close matches, while the machine and panel differed more on Espresso #5."; Tech_Frontiers_Espresso_080310[i++] = new Array("","Predictions","","http://msnbcmedia.msn.com/j/msnbc/Components/Photo_StoryLevel/080227/Espresso-3.hmedium.jpg","","", "", "", "", "", "", "Courtesy of Analytical Chemistry", "", "273", "288", "#000000", "", "", "", ""); Tech_Frontiers_Espresso_080310[i-1].body = "Predictions
A follow-up validation of Nestlé’s profile predictions on eight additional espressos again showed close matches with the profiles assembled by a panel of experts (results for four are shown above). Nestlé’s Christian Lindinger said three descriptors — flowery, winey, and cereal — didn’t make the cut because the espresso samples used for the validation weren’t sufficiently different for those sensory attributes. Even so, he said the validation showed that the electronic nose’s predictions held up for different espresso dilutions and varying years of harvest for the coffee beans."; Tech_Frontiers_Espresso_080310[i++] = new Array("","Nose to \"nose\"","","http://msnbcmedia.msn.com/j/msnbc/Components/Photo_StoryLevel/080227/Espresso-4.hmedium.jpg","","", "", "", "", "", "", "Courtesy of the University of Warwick", "", "273", "404", "#000000", "", "", "", ""); Tech_Frontiers_Espresso_080310[i-1].body = "Nose to \"nose\"
University of Warwick researcher Julian Gardner goes nose to “nose” with an artificial snot-filled sniffer developed by his lab in collaboration with scientists at Leicester University in the United Kingdom. The innovation may help electronic devices become more human-like in discriminating odors. Although the nose knows bananas from milk, it perhaps has some work to do before it can tell jasmine from a more rose-like aroma sniffed by some expert tasters in specialty espressos."; // END editorial data