Researchers from the University of Cambridge have trained a robot chef to taste the food at various stages of the chewing process to determine whether it is sufficiently seasoned – a process that is similar to what humans do. The goal is to have a machine that not only prepares pizza or burgers, but also produces the best meal possible based on human taster feedback.
The robot is already programmed to prepare omelets. Everyone’s tastes differ; some prefer spicy foods, while others prefer sweets. Researchers created a new type of machine learning algorithm that provided the robot with feedback from human samplers, allowing it to improve its product over time, tweaking its methods and eventually whipping up an omelet that tasted great.
Researchers have now collaborated with domestic appliance manufacturer Beko to create a new and improved version of the robot chef, complete with its own taste-testing capability.
We notice a change in texture and taste when we chew our food. During chewing, specific changes occur in the dish’s ingredients. For example, in the height of summer, biting into a fresh tomato releases juices, and as we chew, releasing both saliva and digestive enzymes, our perception of the tomato’s flavor changes. Furthermore, saliva produced during chewing aids in the transport of chemical compounds in food to taste receptors, primarily on the tongue, and signals from taste receptors are transmitted to the brain, where it is determined whether something tastes good or bad.
The researchers attached a conductance probe, which acts as a salinity sensor, to a robot arm to simulate a more realistic human chewing and tasting process in their robot chef. The robot chef tasted nine different variations of a simple dish of scrambled eggs and tomatoes, varying the number of tomatoes and the amount of salt in each dish during testing. The robot used the probe to taste the dishes in a grid-like pattern, returning a reading in just a few seconds.
To simulate the change in texture caused by chewing, the team blended the egg mixture several times before having the robot test the dish again. The robot’s different readings at various points of ‘chewing’ enabled it to produce taste maps of each dish. The results demonstrated a significant improvement in robots’ ability to assess saltiness over other electronic tasting methods, which are frequently time-consuming and only provide a single reading.
“When a robot is learning how to cook, like any other cook, it needs indications of how well it did. We want the robots to understand the concept of taste, which will make them better cooks. In our experiment, the robot can ‘see’ the difference in the food as it’s chewed, which improves its ability to taste.”
co-author Dr. Arsen Abdulali, from the Department of Engineering
The researchers hope to improve the robot chef in the future so that it can taste different types of food and improve sensing capabilities so that it can taste sweet or oily food.