At the end of four years of joint research, the American spice company McCormick andIBM Together they developed a unique artificial intelligence system that helps food engineers and product developers in the food industry create new taste experiences easily
At the end of four years of joint research, the American spice company McCormick andIBM Together they developed a unique artificial intelligence system that helps food engineers and product developers in the food industry create new taste experiences easily.
Creating taste experiences is a challenging task, and the science of how humans experience taste is not fully understood. The well-known claim that four separate taste zones can be mapped on the tongue - sour, sweet, bitter and salty - has been disproved. Most scientists today agree that there is at least one more taste - umami (umami), and the experience of taste is also related to smell and appearance. In addition, scientists believe that the experience of taste occurs simultaneously in different areas on the tongue, and each taste receptor on the tongue is able to detect all five tastes. Also, a genetic component also contributes to the taste experience, which explains why some people like coriander and others dislike it.
Designing new taste experiences is an art that requires a lot of skill and experience. Product developers have a huge number of ingredients to choose from, and they need to determine not only the right combination of ingredients but also the dosages and ratios between them to meet specific requirements. They gain expertise over many years of practical work in processing data and many variables, combining components and raw materials to create formulas and examples, running experiments, consumer testing and learning from the results. The development process consumes a lot of time and resources, and is usually carried out under a tight time frame of launching the product before the competition.
This is where the relative advantages of artificial intelligence come into play, which has the power to simplify and optimize the process. For example, when there is an urgency to bring a product to market that combines vanilla and chocolate flavors, developers may use common vanilla (or chocolate) formulas to shorten development time and speed up the product's time to market. Artificial intelligence technology will be able to explore for the developers, very quickly, a wide range of additional options and locate a less common vanilla formula as an alternative to the routine and familiar formula.
McCormick's first products developed with the help of artificial intelligence will reach grocery shelves in the US in the second half of this year. Based on the promising results so far, the company plans to integrate the system in more than 20 laboratories in 14 countries, encompassing over 500 product developers and taste experiences. The system could help product developers find taste experiences they weren't aware of enough, speed up the time the products arrive on the market and achieve better rating results among consumer tests.
A blog post by the researcher who developed the system
More of the topic in Hayadan:
