Givaudan harnesses AI to optimize food and flavor formulation

19 Feb 2021 --- Givaudan Taste & Wellbeing has launched its Advanced Tools for Modelling (ATOM), which uses artificial intelligence (AI) to optimize food and flavor formulation and facilitate co-creation and collaboration with customers.

These pioneering tools open new doors to creative development and can cut the time to market for new products dramatically, says Givaudan.

Speaking to FoodIngredientsFirst, Rahul Siva, research investigator for taste and well-being at Givaudan, says he expects to see more AI technology being used across F&B sectors in the months ahead.

“For the past two decades, Givaudan has worked in the digital technology space and more recently in big data and AI. The advancements made in technology and knowledge are helping to drive precisely targeted product development, adding value for our customers and consumers.”

Givaudan’s ATOM uses AI to identify positive and negative flavor drivers and explore ingredient synergies. Personalizing taste
ATOM helps with precision, speed, and formula optimization as well as personalized flavors, he remarks. “It has made the process of flavor creation less about trial and error and more systematic, data-driven and intelligent.”

ATOM will be used to optimize flavor formulas to be consumer-preferred, and to make food taste better while being healthy at the same time, adds Siva.

Givaudan works across all categories in food and beverage, so the opportunities are vast, he asserts. 

“The ATOM technology is being integrated into our ecosystem of digital tools: today, it’s already possible for the AI in ATOM to talk to digital tools that allow us to personalize flavor based on individual preferences. We plan to build on these capabilities in the future.”

Identifying flavor drivers
ATOM is the latest addition to an ecosystem of digital and AI tools that allow Givaudan teams in all regions of the world to streamline the end-to-end creation process from conception to rollout. 

Building on over two decades of research, ATOM uses AI and data science techniques to minimize trial and error in the process. The tools identify positive and negative flavor drivers and explore ingredient synergies, to generate new options and insights aligned with consumer preferences. 

The results are then displayed in graphically rich and interactive dashboards that allow Givaudan to co-create with its customers, enhancing creativity and delivering game-changing new food experiences. 

“ATOM strikes the right balance between AI and human intuition, complementing the work of our expert flavorists and developers,” adds Fabio Campanile, Givaudan’s head of global science and technology, for taste and well-being.

“Wide-eyed thinking”
Tools such as ATOM enable Givaudan to help its customers to go beyond consumer expectations, using insight, collaboration and innovation.Givaudan’s digital strategy and approach is broad and comprehensive, ranging from gaining detailed consumer insights to product optimization.

“Implementing what we call ‘wide-eyed thinking’ allows us to use these insights to leverage our curiosity and deep ingredient knowledge to accelerate new product development and create the food experiences of the future,” Campanile comments.

Early projects using the tools have been highly successful. 

ATOM was recently used in a project to reduce salt in cheese snacks. Narrowing down the perfect blend of ingredients would usually take extensive trial and error. Still, the team quickly identified the ideal recipe, delivering a 33 percent reduction in salt, from a much smaller range of options predicted by ATOM. 

In blind taste testing, the reduced salt recipe scored as highly as the original full-salt snack, making it “just as tasty,” with 33 percent less salt. The process has also proved highly successful for sugar reduction, vanillin replacement, and meat alternative projects. 

By Elizabeth Green