[ad_1]
In October, IBM Research unveiled Philyra AI, a tool to accelerate the creation of new fragrances for the fragrance industry. "It's a system that uses new and advanced machine learning algorithms to sift through hundreds of thousands of formulas and thousands of raw materials," said Dr. Richard Goodwin, Principal Investigator and Research Group Manager. about IBM's IT creativity last October. novel motifs and combinations ".
For example, the system can take into account alternative raw material supplements and replace them with formulas if the original ingredients are not available, adjust the formulas according to the desired reaction of the user, such as "usability" or "gender balance" and even change the recommended dosage. levels based on usage patterns.
The transition from blending fragrances to blending spices is not great and, with the McCormick partnership, the Philyra AI has been easily reconverted for this new task. In fact, Philyra AI was used for the first time to help Symrise produce a new pair of fragrances for O Boticário, a global beauty company, which will be released later this year.
Currently, McCormick relies on a Stage Gate 5-step system to create flavor profiles for its various products – ideation, development, testing, scaling, marketing and maintenance – with most of the effort going to the first two stages. For example, if McCormick's R & D department is working on a new Tuscan chicken recipe, researchers will first search the database for recipes for Tuscan chicken recipes and use them as a starting point before to adjust the flavors and begin the iterative process of focusing in the final product. These tests may involve hundreds of trials lasting 6 to 18 months.
The Philyra algorithm significantly reduces this test time because, well, computers. Artificial intelligence not only accesses the existing database of aroma and spice formulas, it also integrates sales and marketing data, which allows 39, business to adapt its revenue to specific regions and customer demographics. In this way, kosher, halal, vegan or low sodium versions will taste as close as possible to the originals.
"This machine system would allow us to really search for one that has such a broad hedonic acceptance that it could become the next icon," said Engadget, Dr. Hamed Faridi, chief scientist for McCormick. In addition, this will enable the more than 500 members of the company's R & D department located around the world to integrate their diverse skills and expertise into a singular development effort. McCormick hopes to create the next iconic food item: the next Philadelphia Cream Cheese, Pringles Chips, Tootsie Roll or A1 Steak Sauce.
"Looking for an icon looks like looking for a needle in a haystack," Faridi explained. But this AI should help cut through the ball. "If I'm a developer and want to develop something that, I think, needs a cheese flavor, I have two or three of my favorite cheese flavors that have worked in the past. I'm trying to use them first [because of their previous success]"However, AI does not have such prejudices and will rather weigh the different cheese flavors present in McCormick's database in order to find the one that is" better "hedonically.
McCormick has already leveraged artificial intelligence to develop three new recipe mix flavors, Tuscan chicken, Bourbon pork fillet and New Orleans sausage, which she plans to publish from here the middle of the year. The company plans to expand access to the algorithmic system to all of its research and development workforce by 2021.
Source link
