This browser is not actively supported anymore. For the best passle experience, we strongly recommend you upgrade your browser.
| less than a minute read

Robot chef learns to make salad by watching humans

Cambridge researchers have developed a robot capable of watching a cookery video, recognising which of several simple salads is being made, and then making the same salad itself using its robotic arm. The robot even demonstrated the impressive ability to watch a new, unfamiliar recipe being made, determine the required steps, and add the new recipe to its programmed cookbook. Perhaps one day robot chefs will be a familiar sight! At the very least, surely it won't be long until the robot is making up its own recipes (presumably with mixed success) rather than just copying them.

It's notable that the researchers trained their aspiring restaurateur using "a publicly available neural network". The pace of innovation in AI research is tremendously fast, as shown in Marks & Clerk's recent report on AI-related patent activity, and new tools are becoming widely available all the time. Whenever the pace of progress is so quick, it's vital to consider patenting new ideas as soon as possible. We would be delighted to help protect any further robotic culinary endeavours!

By correctly identifying the ingredients and the actions of the human chef, the robot could determine which of the recipes was being prepared. The robot could infer that if the human demonstrator was holding a knife in one hand and a carrot in the other, the carrot would then get chopped up. Of the 16 videos it watched, the robot recognised the correct recipe 93% of the time, even though it only detected 83% of the human chef’s actions. The robot was also able to detect that slight variations in a recipe, such as making a double portion or normal human error, were variations and not a new recipe. The robot also correctly recognised the demonstration of a new, ninth salad, added it to its cookbook and made it.