Imagine that you are about to take your first bite into a brownie ice cream sundae, or your mom’s homemade special lasagna. I bet the memory of your favorite foods is making your taste buds tingle. Right. This. Second.
Now picture a computer “tasting” the difference between an apple and apple pie. Is a computer algorithm really capable of mimicking our taste system?
Before we learn about this new advancement, let’s first explore the details of how we are able to taste. What if I told you your taste cells are really only capable of identifying five different tastes? Certainly, something in addition to our gustatory system, or the sensory system that detects taste, allows us to taste the difference between a double-bacon cheeseburger and a steak, but what is it?
HOW DO WE TASTE?
The human tongue has about 10,000 taste buds, or clusters of 50 to 150 cells, that detect tastes. Each taste cell is designed to detect only one of the five tastes human are capable of experiencing: salty, sour, sweet, bitter, and umami (also known as the meaty, savory flavor). We sense these tastes when certain ions or molecules bind to ion channels or protein receptors, sending signals through several cranial nerves to the thalamus (a relay station for information), and finally to the gustatory cortex of the brain.
While taste cells are only capable of distinguishing between five different tastes, our olfactory epithelium, or cells that line the back of our nasal cavities and transmit odor information to the brain, are capable of detecting more than 10,000 different odors.
It is a combination of odor and taste information that produces an overall flavor. So when you’re eating that rich, velvety, dark chocolate cake versus a generic candy bar, thank your olfactory system for your ability to taste the difference in both sweets.
CAN A COMPUTER REALLY TASTE?
After learning how humans use both taste and smell to detect different flavors, how can it be made possible for computers to detect different tastes? Well, scientists are building a kind of electronic tongue that is really a biosensor that simulates the tasting abilities of the human tongue. It has an array of chemical sensors that can interact with salts, sugars, bitter molecules, and other chemicals in a liquid sample and can send electrical signals to a computer for mathematical and statistical analysis to extract taste information. For example, one type of electronic tongue uses fatty membranes as sensors: salts, sugars, and other taste compounds interact with the membrane and change its electrical composition.
For now, the electronic tongue can only be calibrated using known taste profiles in order to determine the difference between new flavors. As a result, this device can’t necessarily identify a food item based on its taste, but it can be used to sense the change in bitterness and acidity of sour milk over time (gross!), which is a very important advancement in the food industry. The real potential for this technology will come into play when the pharmaceutical industry uses the electronic tongue to improve the flavor profiles of chemicals and drugs.
So who knows, maybe soon we’ll have our choice of chocolate or bacon flavored cold medication instead of the boring grape and cherry flavors we’re so used to.
S. V. Litvinenko et al., Might silicon surface be used for electronic tongue application?, ACS Appl. Mater. Interfaces, 6(21):18440–18444, 2014.