Artificial Intelligence as rapid prototyping tool in flavours

Favoritar este trabalho
Como citar esse trabalho?
Detalhes
  • Tipo de apresentação: Oral
  • Eixo temático: Métodos sensoriais e/ou Emergentes
  • Palavras chaves: Artificial Intelligence; VAS; ATOM; Rapid Prototyping; drivers-of-liking;
  • 1 Givaudan / Consumer Sensory Insights Mexico Flavours
  • 2 Sensory / Private Industry / Givaudan de México
  • 3 Givaudan de México

Artificial Intelligence as rapid prototyping tool in flavours

Irery del Rocio Sanchez Martinez

Givaudan / Consumer Sensory Insights Mexico Flavours

Resumo

Consumers are at the heart of our flavours business. In the pursuit of fully understanding consumer desires, Givaudan has developed different approaches to delivering quantitative and qualitative research, aided by our own advanced Smartools technologies. 2

Smartools enable us to collect consumer preference data effectively and translate consumer information into actionable for flavour and taste creation. VAS® (Virtual Aroma Synthesizer) is a Givaudan’s proprietary technology that is designed to create a flavor / fragrance in real time with customers/consumers and understand consumer preferences by precisely blending ingredients, keys or accords. Givaudan’s proprietary software allows you to create any blend, smell the aromas in real time, save the aroma profiles you liked and finally convert the aroma into formula using smell to taste algorithm. 3
ATOM® (Advanced Tools for Modeling) is an artificial intelligence recently developed by Givaudan. It uses Artificial Intelligence algorithms to generate advanced DOE (Design of experiments), collect and analyze sensory data , identifies Driver-of-Liking , offer personalized flavors to every user, as well as recommend market products in real-time with advanced statistical modeling technology. 4 In this research, VAS® & ATOM® Technologies were used for rapid prototyping of strawberry flavours in a beverage. A quantitative consumer test was carried out with six different strawberry aromatic keys and 120 consumers. Consumer data were captured in public arena in order to get the best or “ideal, strawberry flavour profile for consumers in Mexico. Based on the consumer data, ATOM, the artificial intelligence generated three best strawberry profiles with a high predicted estimation of overall liking score; one of them got 9 points in a 9 point overall liking scale. We decipher the drivers-of-liking for this strawberry flavor combination, the positives and the non-positive ones. In the end we derived rapid prototyping within five days, for strawberry flavour using the integration of artificial intelligence and consumer responses.

Compartilhe suas ideias ou dúvidas com os autores!

Sabia que o maior estímulo no desenvolvimento científico e cultural é a curiosidade? Deixe seus questionamentos ou sugestões para o autor!

Faça login para interagir

Tem uma dúvida ou sugestão? Compartilhe seu feedback com os autores!