Cambridge University trains robot to 'taste' as it cooks eggs

The robot uses a probe to "taste" the food as it cooks

A robot has been trained to "taste" food at different stages of the chewing process to check if it is salty enough, in a way similar to humans.

Researchers from the University of Cambridge said it could help in automated food preparation.

The robot, which had already been trained to make omelettes, tasted a dish of scrambled eggs and tomatoes.

The paper's co-author Dr Arsen Abdulali, said it could make the robots "better cooks".

According to the research, the robot tasted nine different variations of scrambled eggs and tomatoes at three different stages of the chewing process.

In order to imitate the change in texture caused by chewing, the team put the egg mixture in a blender and had the robot test the dish.

To imitate tasting in their robot chef, researchers attached a probe - which acts like a saltiness sensor - to a robot arm.

Using the probe, the robot "tasted" the dishes in a grid-like fashion, returning a reading in just a few seconds.

It then produced taste maps of the different dishes.

The researchers found the "taste as you go" approach improved the robot's ability to quickly and accurately assess the saltiness of the dish.

• Anglo-Saxon kings mostly vegetarian, study finds

• 'Stolen' Darwin notebooks mysteriously returned

• Blue plaque to honour 'Big Bang' scientist

Dr Abdulali said: "We wanted to replicate a more realistic process of chewing and tasting in a robotic system, which should result in a tastier end-product.

"When a robot is learning how to cook, like any other cook, it needs indications of how well it did.

"In our experiment, the robot can 'see' the difference in the food as it's chewed, which improves its ability to taste."

Grzegorz Sochacki, from the university's Department of Engineering, who co-wrote the paper, said: "If robots are to be used for certain aspects of food preparation, it's important that they are able to 'taste' what they're cooking."

The findings were published in the journal Frontiers in Robotics & AI.