How AI and digital help us innovate faster and smarter
Published:
From the launch of the very first digital tool to the recent formulation of the world’s first green carbon detergent, Unilever R&D experts have been using digital technologies to remain at the forefront of superior innovation for years. And we’re just getting started.
As consumer expectations of our products and brands continue to evolve, we’re working to meet their demands against a backdrop of supply chain disruption, cost inflation and the impact of an escalating climate crisis.
Digitalisation allows us to respond to all these challenges by unlocking radical new levels of collaboration, capability and consumer centricity.
Unilever R&D teams have been innovating for over 100 years to create superior brands and products. Today, our experts are keeping us at the forefront of that process by applying the latest digital technologies to our ‘discover–design–deploy’ approach.
Changing how we discover
We’re discovering new ingredients and new science in a way that was never imaginable before digital.
We’re making huge discoveries about human biology and how the body interacts with our microbiome which is only possible because of access to large-scale data, machine learning and high-performance computing, making connections the human brain would never even be able to achieve by itself.
We’re on the cusp of industrialising biotechnology together with our partners to commercialise plant-based, sustainable palm oil alternatives for use in everyday cleaning and personal care products.
And bringing together the very best in technology, data and people, we’re identifying and predicting consumer trends before they even happen.
Progressing how we design
To turn people’s needs into sustainable, cutting-edge innovations, we map, model and experiment virtually, designing and simulating process flows before these innovations go on to be manufactured at scale in our factories globally.
This way of designing products ‘in-silico’ (using computational models) allows us to understand the effect and interaction of molecular compounds – to determine how our formulations evolve, react and even biodegrade – faster and more effectively than ever before with virtual simulations.
Thanks to in-silico, we can test millions of recipe combinations in seconds which helps us design products to suit regional taste preferences. With Hellmann’s, for example, we make gravy and coronation variants of our mayonnaise for the UK market, and sriracha and lime sauces for US consumers.
It’s also at the heart of our work on non-animal testing. We don’t agree that animal testing is necessary to assure the safety of ourproducts or the ingredients in them, and support calls for a worldwide animal testing ban on cosmetics by 2023, working with government authorities, NGOs and our suppliers across the world to increase the use of non-animal approaches for regulatory compliance purposes. For example, using AI, we can predict the response of the biological process when the skin is exposed to certain chemicals or ingredients.
Transforming how we deploy
We’re combining R&D and Supply Chain data with manufacturing simulations to optimise how we deploy our products to 3.4 billion daily consumers worldwide through our global supplier base and manufacturing presence. Being able to respond to real-time trends and market changes more quickly, we can make our innovations more effective and desirable, scaling them faster than ever before.
That said, it’s not enough just to design better-performing products. We need to innovate in a responsible, future-fit and sustainable way to keep on serving the needs of our consumers while protecting the planet for future generations of consumers.
As Alberto Prado, Unilever R&D’s Head of Digital & Partnerships, says: “I firmly believe that science and technology will play a central role in helping innovate for a better future. The combination of digital and leading-edge science is empowering our teams to progress their fields not by years, but by decades.”
Improving operational efficiency
We’re using AI to help identify alternative ingredients that can strengthen the resilience of our supply chain, making our formulations more sustainable and cost-efficient, and simplifying them by reducing the number of ingredients without impacting a product’s quality or effectiveness.
By optimising our design and manufacturing processes through virtual simulations, automated workflows and data-led decision making, we’re unleashing new levels of efficiency in our operations which is in turn freeing up our experts to focus on delivering impactful innovations.
“We have been scaling the use of digitaltechnologies to work smarter for years,” says Alberto. “The game-changing digital tools we use today strengthen our insights and capabilities to power next-level innovation, product reformulations, scientific discovery and portfolio simplification. And we have just scratched the surface of what’s possible.
“Digital is helping us become much more effective at satisfying the needs of consumers, but also those of the planet by creating increasingly sustainable products.”
The secret to helping dry skin heal itself
Through AI-powered data analysis, we researched the microbiome – the 100 trillion+ microbes in, on and around our body – and its role in stimulating our immune system and keeping the skin healthy. We discovered how the skin microbiome population and skin ceramides can both be managed with products to help resolve issues in skin quality and hydration.
To understand the interactions between the microbiome and external forces, we had to dive into 12 terabytes of data and extract unique insights, much of which has gone already into everyday products such as Dove, Pond’s and Vaseline to create revolutionary, patented technology, delivering healthy-looking skin for consumers.
The deodorant powered by artificial intelligence
Axe A.I. Body Spray was developed using 46 terabytes of data, 6,000 ingredients and 3.5 million potential fragrance combinations to create one unique scent.
The new limited-edition fragrance was developed to tap into Gen Z’s fascination with the world of tech and crypto, and drive excitement into the deodorants category.
We created the product in collaboration with Swiss fragrance company Firmenich, who combined modelling, formula prediction and optimisation with Axe’s consumer trends knowledge and expertise to come up with the unique scent combination.
The advanced tech at the heart of a beauty revolution
Over recent years, we’ve developed more than 40 technologies for our Prestige Beauty business, which is growing fast, with a turnover of €1.2 billion in 2022. One of our most recent innovations is Hourglass Red 0. This is the first vegan alternative to the most commonly used red pigment in colour cosmetics – red carmine, which is derived from female beetles and can take 1,000 beetles for one single lipstick.
We reformulated the whole thing using computational models. It took our scientists three years to develop but we’re planning to make the process open-source so others can apply the same technology. It’s a perfect fit with Hourglass’s ongoing commitment to animal welfare.
The machine learning behind clothes care
Comfort Ultimate Care fabric conditioners contain patented technology that protects garment fibres so they retain their as-new colour and shape for up to 20 washes. But it’s a tricky business to work out what’s the right balance of active ingredients to achieve that in different parts of the world when you have to consider things like variances in water hardness and quality.
So the Comfort R&D team designed experiments using machine learning to work out the best product formula based on the local conditions in each market. If they ever need to tweak a formula, they simply re-optimise to new conditions and can change the formula within hours.
The way we improve the nutritional quality of our products
To help people enjoy healthier diets, we’re continuously improving the nutritional quality of our food and beverage portfolio globally, by reducing sugar, salt and saturated fat, and also fortifying our products with micronutrients.
For more than a decade, we have used our own nutrient profiling model to assess the levels of nutrients and healthiness of our products. But there are many other models in use worldwide, which often arrive at different conclusions. So we now use our model alongside six others.
Our digital capability was crucial to achieve this in a short time. We were able to assess more than 600,000 pieces of data about the ingredients and recipes of all our products globally.
The smart cabinets keeping ice cream lovers happy
In our out-of-home ice cream business, we’re using more and more data from the millions of cabinets we have across the world. This allows us to monitor the quality of the products delivered and also ensure the store offers the optimal range of products that we know its local shoppers want.
These ‘smart freezers’ identify when the stock of a particular product in the cabinet is running low and send a message to the store team that suggests a quantity to reorder. These are then dispatched automatically, drastically reducing the chance of the store running out.
When research showed that shoppers want a toilet and bathroom cleaner that kills germs more effortlessly in hardest-to-reach places, Domestos came up with its biggest innovation in a decade: one that’s set to continue the brand’s commitment to leading the growth of the category.
We’re a founder member of the International Collaboration on Cosmetics Safety, working with NGOs and industry experts to accelerate global use of animal-free safety science.