Robot Chef That Cooks Omelettes Learns to Assess Taste of Dish During Chewing Process

A team of researchers trained a robot chef to assess the taste of a dish at various stages of the chewing process, mimicking a similar process in humans.

As indicated in a Science Daily report, the results could help develop automated or food preparation by assisting robots to learn about good and bad taste, "making them better cooks."

@twitter|https://twitter.com/PopSci/status/1522035651943751680?ref_src=twsrc%5Etfw@

When food is chewed, one can notice a change in both taste and texture. For instance, biting into a fresh tomato at the height of the summer season will release juices, and as humans chew, releasing both saliva and digestive enzymes, the perception of the flavor of tomato will change.

Essentially, the robot chef, who has been trained to make omelets based on the feedback of a human taster, tasted nine different variations of a simple dish like scrambled eggs and tomatoes at three stages of the chewing process and produced taste maps of the different dishes.

'Taste as You Go'

Researchers from the University of Cambridge discovered that this approach, called "taste as you go," substantially improved the ability of the robot to swiftly and precisely assess a dish's saltiness over other electronic tasting technologies, which only test one homogenized specimen. The results came out in the Frontiers in Robotics & AI journal.

The perception of taste is a complicated princess in humans that has developed over millions of years. Specifically, the appearance, texture, temperature, and smell of food are all affecting how humans perceive taste, the saliva produced while chewing helps carry chemical elements in food to tasted receptors, mostly on the tongue, and the indications from taste receptors are passed to the brain.

Once the brain becomes aware of the flavor, one then decides whether he is enjoying the food or not. Essentially, the taste is not highly individual, as well. Some people are fond of spicy foods, while others have the so-called "sweet tooth."

Whether he is an amateur or a professional, a good cook depends on his sense of taste and can balance the various flavors within a single dish to make a well-rounded final product.

Training Robots

The study investigators are members of the Bio-Inspired Robotics Laboratory of Cambridge, which is run by Professor Fumiya Iida of the Department of Engineering, and focus on training robots to perform the "last-minute meter problems" that humans consider easy but difficult for robots.

Cooking is one of these tasks. Earlier tests, in particular, where their "robot chef" has produced a passable omelet through feedback from human testers.

According to the paper's first author, Grzegorz Sochacki from the Department of Engineering of Cambridge, "We need something cheap," tiny and quick to add to the robot to perform the tasting.

This needed to be cheap enough to use in a kitchen, tiny enough for a robot, and quick enough to use while cooking, added Sochacki.

Mimicking the Human Process of Chewing and Tasting

To mimic the human process of chewing and tasting on their robot chef, the study authors attached a conductance probe, which acts as a "salinity sensor to a robot arm."

A similar Popular Science report said they prepared tomatoes and scrambled eggs, varying the tomatoes, number, and salt amount in every dish.

Using the probe, the robot performed the tasting of the dishes in a grid-like manner, returning a result in just a few seconds.

Meanwhile, to mimic the change in texture resulting from chewing, the team then put the egg mixture using a blender and had the robot test the dish once more. The different outcomes at different " chewing " points yielded every word's taste maps.

Information about this robot chef is shown on Randomonic's YouTube video below:

Check out more news and information on Robots in Science Times.