Unilever is leveraging artificial intelligence (AI) to enhance its product development process, particularly in the food segment, which includes well-known brands such as Hellmann’s and Knorr.
By integrating AI technologies, the company has been able to not only speed up the creation of new products but also refine their predictability in terms of consumer reception and production efficiencies.
Manfred Aben, Unilever’s global vice president of science and technology for nutrition and ice cream, is at the forefront of these advancements.
With nearly three decades at the company, Aben has witnessed the evolution of AI from simple rule-based systems to complex, data-driven models that now play a critical role in product development. Initially used for assessing the shelf life of products, AI’s role has expanded significantly under his guidance.
Unilever’s Product Development: A Blend of Tradition and Innovation
The traditional product development cycle at Unilever starts in the professional kitchen, where chefs experiment with new ingredient combinations. Successful recipes are then tested with consumers to select the most promising ones for mass production. Although this process still requires human involvement and physical testing in pilot plants and factories, AI has introduced a higher degree of predictability and efficiency. Aben explains that AI allows for fewer physical trials by accurately simulating the effects of different settings, ingredient dosages, and timings.
One of the key objectives for Unilever is to make its nutritional products “holistically superior,” which involves ensuring they are healthy, environmentally friendly, and affordable. This goal presents a complex optimization challenge that AI helps to simplify by providing powerful tools throughout the development cycle, from consumer insights to the final product.
AI in Action: Case Studies of Unilever’s Innovations
A notable innovation from Unilever that utilized AI is the development of the Knorr Zero Salt Bouillon Cubes. Addressing the challenge of reducing salt without compromising taste, structure, or production performance, AI modeling predicted the outcomes of using different ingredients, thus minimizing reliance on trial and error.
Additionally, AI was instrumental in redesigning packaging for Hellmann’s vegan mayonnaise to better handle its unique consistency compared to traditional mayonnaise. By simulating different materials and coatings, Unilever could optimize the packaging design more quickly and efficiently.
Aben also highlighted how AI aids in understanding consumer preferences, such as predicting whether a product might be perceived as too sweet or savory. This capability is crucial for rapidly responding to market trends and refining existing products. Despite the technological advancements, the final validation with actual consumer testing remains indispensable, ensuring that the products meet real-world expectations.
Overall, the integration of AI into Unilever’s product development has not only shortened the innovation cycle from months to weeks or days but has also allowed the company to increase the volume and variety of new products it introduces to the market. This shift towards more rapid and data-driven innovation is in response to growing consumer demands for new and improved food options.
Related News:
Featured Image courtesy of Poulssen/Getty Images