Artificial Intelligence is making more and more of an impact on our daily lives, and making more headlines than ever. But for some time now, AI has been part of the process of flavors and fragrances development, with sometimes surprising results.
The Demand for the New
The fragrance industry is driven partly by ongoing demand for the timeless fragrances that become a person’s signature scent, and partly by demand for new and unusual fragrances that intrigue new customers.
In recent decades, perfume has become an essential part of the personal grooming process for more customers around the world. The fragrance industry enjoys a projected growth rate of 5.9% through 2030, reaching over $80 billion USD, with the greatest demand in Europe.
AI in Scent Identification
In the past, fragrance formulas were heavily guarded secrets, shrouded in mystery for decades. Since the 1980s, sensory technologies have become more and more accurate, making it possible to recognize, and thus reverse-engineer, fragrances. With the possibility of creating affordable duplicates, fragrance brands, like fashion, rely on reputation, quality, packaging, and brand name prestige to continue to differentiate their products.
Scent identification technologies, powered by artificial intelligence, are faster and more accurate at identifying odor compounds than humans are. Google’s Odor Map can also analyze scent molecules and predict how they will activate scent receptors in different animals, providing unprecedented insight into the biology of fragrances.
AI scent identification technologies are being used in a variety of ways, including:
- Developing novel fragrance combinations. Scent is strongly tied to memory, so humans often associate and combine fragrances in similar ways. AI can be used to suggest fragrance compounds and combinations that are highly unusual, without the typical associations. While these novel combinations aren’t always appealing, they can stimulate creativity and inspiration for perfumers.
- Find scent duplicates for rare ingredients. Many of the most important fragrance ingredients are rare and hard-to-find, or are hard to produce in a sustainable way. Highly accurate identification allows manufacturers to find more available or more sustainable alternatives with the exact same chemical structure.
- Identify irritants and allergens in complex formulas. For manufacturers of products that use fragrances, like shampoos, laundry detergents, cosmetics, etc., it can be hard to identify the exact chemical or compound that is acting as an irritant or allergen. Scent de-formulation can allow a company to identify and isolate specific compounds for further testing.
AI in the Flavor Industry
Like fragrances, flavors are subtle, complex, and crucial to the consumer. It is playing an important role in the future of flavor discovery and customization. Some of the companies using AI in flavor development include:
- Firmenich. In 2020, Switzerland’s Firmenich announced the world’s first AI-created flavor. Using AI along with the company’s existing flavor and fragrance material database, it developed the flavor of light grilled beef for use in plant-based proteins.
- Foodpairing. Belgium’s Foodpairing uses gas chromatography and mass spectrometry to build a database of ingredients and flavors, and then uses AI algorithms to develop novel flavors for foods and cocktails.
- McCormick’s ONE. American spice company McCormick partnered with IBM to create specialty spice packets and recipe mixes for complex flavors like Tuscan chicken, New Orleans sausage, and bourbon pork. The algorithm can also adjust flavor compounds to align with marketing strategy and regional preferences.
Grapefrute salutes these innovative companies using AI to make a more fragrant, flavorful, sustainable world. To be part of the next generation of Fragrances and Flavors innovators, you need the right talent and the right team, so contact us today to learn more.