Using AI to develop data-driven scratch cooking aids [Interview]

At this year’s Fi Europe, Marcelo Camilo de Oliveira, head of food technology for cooking aids and meals at Unilever Netherlands’ global research and development team, will be sharing his insights and experiences of using AI in product development.

He will be a key participant in an expert panel entitled ‘Practical applications of AI and the role of tech-driven innovation in the food industry’. “I’ll be bringing the product development point of view to these discussions,” he says. “There is a lot to be explored in this space.” 

Predictive modelling in product development

One of the major challenges for product developers is time. It can take years – and millions of euros – to create products that meet consumer expectations as well as regulatory requirements. Being able to predict what a final recipe will look like, taste like, and how it behaves during manufacturing can help manufacturers achieve significant cost and time savings.

“There are so many dots to connect within product development, and everything is moving so quickly,” explains de Oliveira. “Being able to connect all these dots efficiently would be really beneficial – and this is where AI can help us.”

De Oliveira’s team is focused on developing technologies and innovations for scratch cooking. “If you’re in the kitchen, you might need ingredients or products that can help you make your food taste great, or give it great texture,” he says. “We try to unlock these attributes through food tech and food science. I currently head the team that is coming up with innovations that can help with food preparation and make daily meals taste always great.”

The team is currently applying AI to identify possible areas of improvement here. Data has been gathered to build up predictive modelling capabilities in certain product attribute areas such as colour and texture. The aim is to go beyond human predictive capabilities, to speed up the development process.

De Oliveira will be sharing these experiences at Fi Europe. “It has been a journey – there have been a number of challenges we’ve had to address,” he says. “These include getting the right amount of data, ensuring the quality of data, and then managing all this data.”

Complementing human expertise with effective AI

De Oliveira will also be discussing many of the lessons learned along the way. “Using AI does not mean to say that we no longer need product developers,” he says. “We need their knowledge to judge and evaluate the options popping up in front of them. In our experience, successful product development is about getting the right combination of human expertise and effective AI.”

Another lesson has been that data is not the be all and end all. The temptation can often be to continually collect more data, to build more powerful predictive AI tools.

“The danger is that data becomes a never-ending conversation,” says de Oliveira. “Being effective is not only a question of data creation, but of knowing how this data can lead to new business opportunities. Too much focus on the data can prevent you from applying perspective to other areas of the business.”

For de Oliveria, it is a question of balance. Experts are still needed to input, select, and apply their expertise throughout the product development phase. AI can speed up processes and point teams in the right direction, but in-house knowledge is still essential.

“You can only take advantage of predictive models if you have the right people behind them,” he says. “This is how we see the future – AI working with experts, as if both are in front of the computer. This is how great products will be delivered in the future.”

This AI panel discussion will take place during the Future of Nutrition Summit and will feature other leading innovators in the field. A key goal of the session is to explore cutting-edge AI and tech-driven innovations that are shaping the future of food.

In addition to Unilever’s Marcelo Camilo de Oliveira, NotCo’s CEO Matias Muchnick will be discussing how AI is revolutionising ingredient solutions and food product design, while Francis Taloen of Kline & Company will outline some of the real-world applications of tech-driven strategies in food.