PepsiCo, Danone & Nestlé: how AI is powering F&B growth

AI in food: summary

• AI in F&B boosts growth with predictive analytics to help forecast demand and reduce waste
• Digital twins and AI simulations improve manufacturing and logistics by cutting costs and preventing disruptions
• AI speeds up product innovation by using data on trends like gut health and personalised nutrition
• Automation improves compliance and traceability, which strengthens food safety and reduces errors
• AI supports sustainability as it can track carbon, optimise energy use and improve sourcing

Predicting the future is usually covered by clairvoyants and analysts, but an increasing number of food and drink manufacturers are now employing artificial intelligence (AI) to do the job.

AI has already rolled up its sleeves and streamlined services in offices and factories, and now the likes of Nestlé, PepsiCo, Danone and Unilever are putting it to work to help in a predictive capacity – to anticipate consumer demand and foresee and mitigate problems before they occur.

From using it to identify the most resilient crops for its products to better planning of warehouse space, AI technology is taking a huge amount of guesswork out of the equation for food and drink manufacturers. In this age of global uncertainty, AI’s ability to analyse data quickly from all areas of an organisation – even when the data is limited – is a useful tool.

“Food manufacturing requires constant trade‑offs between quality, cost, service and sustainability,” explains Christophe Villain, senior global operations and digital transformation leader at Nestlé.

“AI helps make those trade‑offs more transparent and manageable, allowing companies to respond faster to consumer needs without compromising safety or standards.”

Villain, who leads end-to-end operations and digital transformation at the Swiss-based manufacturer, said beyond efficiency, Nestlé was using AI to help it anticipate consumer needs to ensure products were delivered in the right place at the right time.

“This includes using AI and causal models to simulate complex scenarios impacting our end-to-end value chain – for example, how changes in demand, availability of ingredients or logistics constraints could affect product availability or freshness,” he says. “These simulations help operations teams adjust plans before consumers feel the impact.”

Logistical help

AI is also aiding logistics. In the US, PepsiCo is using digital twins (virtual models) and AI-agents to co-design new production and warehousing facilities. The move means the business can test the space for capacity and efficiency before spending huge sums on a build.

Danone and Unilever are among the businesses using AI to suggest the kinds of products its customers will want in the future.

Danone said it had enlisted the help of predictive modelling and AI-enabled research in areas such as gut health to ‘develop products with evolving needs’, while FMCG giant Unilever is doing similar in its beauty and wellness division.

Aside from predictive modelling, AI is being used by food manufacturers to aid areas such as compliance, traceability and quality control, says Stephen Pavlich of Loftware.

The labelling software company uses AI to automate label checks by reviewing ingredient lists, allergen declarations, nutritional information and country-of-origin details against regulatory requirements such as Natasha’s Law and FSMA 204.

“Our work focuses on helping businesses move away from manual, fragmented processes toward connected product identification systems that reduce errors and improve responsiveness,” says Pavlich, noting growing interest in connected packaging strategies that link physical products with digital information and consumer engagement experiences.

Sustainability gains

AI is also doing some heavy lifting in helping to cut carbon emissions. It’s being used across the supply chain to measure, reduce and redesign emissions-intensive activities, says Sam Stark, CEO and founder of Green Project Technologies, with the biggest impact made by combining AI with operational data.

“Unlike traditional rule‑based systems, AI can learn from imperfect and incomplete data, which happens often in food supply chains where supplier data is inconsistent. This makes Scope 3 emissions measurable and actionable much earlier than previously possible.”

As with other areas where AI is used, it is its ability to speed things up in carbon reporting that is its forte.

“AI shifts decarbonisation from reactive reporting to predictive action. By forecasting demand, yields, energy use and emissions trajectories, it enables early decision-making to avoid excess emissions being generated,” concludes Stark.

How the big brands are using AI

PepsiCo

PepsiCo is using AI and digital twin technology to help ‘retool and optimise’ its existing physical footprint in response to rising demand for production and distribution capacity.

The collaboration with Siemens and NVIDIA, which is currently being piloted in the US before being rolled out globally, will use digital twins (virtual models) and AI-agents to co-design new production and warehousing facilities.

PepsiCo chairman and CEO Ramon Laguarta said the business was ‘embedding AI throughout operations’ to help it meet increasing demands from customers.

“The scale and complexity of PepsiCo’s business, from farm to shelf, is massive,” he says.

A Siemens digital twin composer, built on NVDIA Omniverse libraries digital twins and AI agents, will be used to simulate upgrades to its current facilities, allowing the business to test the space before committing to a physical build.

The technology allows every machine, conveyor, pallet route and operator path to be recreated ‘with physics level accuracy’, according to the Pepsi-Cola and Lay’s producer.

AI agents then simulate, test and refine system changes with the ability to identify up to 90 per cent of potential issues before they physically occur. The approach is said to drive faster design cycles – teams reported they were able to validate new configurations that boosted capacity and throughput ‘within weeks’ – and can reduce capital expenditure by up to 15 per cent.

Athina Kanioura, CEO, Latin America, and global chief strategy & transformation officer at PepsiCo, said the digital blueprint was a first for the industry.

“With a unified, AI-powered digital foundation, PepsiCo is building toward a world where every plant and warehouse operates as part of a single, intelligent ecosystem. In this future, our facilities don’t just respond to demand, they anticipate and then adapt to it.”

Danone

AI’s use at the French dairy and water major is focused in three areas: developing new products, making operations more efficient and sustainable, and staff training.

The first – innovation – is arguably its most creative use. Here, Danone has enlisted predictive modelling and AI-enabled research in health and wellness to ‘accelerate science and innovation’ in areas like gut health that will help it develop new products that it says will meet consumers’ ‘evolving needs’.

Efficiencies are less exciting, but essential for a successful business. AI supports Danone’s industrial operations to make plants safer, reliable and resource efficient, it says.

“In practice, we use digital tools that help teams anticipate equipment issues through predictive maintenance and optimise key production parameters, while keeping human expertise in control,” says a spokesperson.

Danone is also one of a growing number of manufacturers automating its product carbon footprinting with sustainability research company and SaaS platform HowGood.

Where AI is making the biggest difference at Danone, however, is in supporting its teams. The business is collaborating with Microsoft to build on the work of its DanSkills programme, which trains teams in how best to use AI. For example, in digital twin technology to simulate and test changes and predictive maintenance to anticipate issues.

“For us, the biggest benefits of AI are people enablement – helping teams on the shop floor, in supply chain and in R&D get better visibility and decision support – and reducing manual admin,” it says.

Using data and AI to cut energy use and wastage and to analyse customer behaviours to better deliver services for them are also major benefits.

Nestlé

AI isn’t considered a support service at Nestlé. The Swiss manufacturing giant regards it as a core capability and “an integral part of how Nestlé operates its global manufacturing and supply chain network”, says Nestlé’s Villain.

Today, AI is used across multiple areas of the business, from ensuring that the demand and supply of its products align to optimising performance in factories.

For example, the AI-based forecasting tools it uses analyse historical sales and other factors to better anticipate consumer demand.

“This reduces the risk of over‑ or under‑production, improves on‑shelf availability and limits waste – directly affecting what consumers find in stores,” says Villain.

At the other end of the supply chain, it has enlisted the help of AI‑supported climate‑resilient agriculture, using advanced data models to select and breed more resilient crop varieties. It is also used to accelerate the development of more sustainable packaging materials.

On the topic of sustainability, AI is helping to make Nestlé more energy efficient by analysing material flows, quality losses and energy consumption. The regular analysis gives teams a more in-depth view of food and energy wastage, says Villain, meaning the business can react quickly to reducing it.

Like PepsiCo, Nestlé has employed digital twins in its factories through a partnership with Accenture Song (built on NVIDIA Omniverse) to test changes virtually. This includes using AI and causal models to simulate complex scenarios, such as how a change in demand, availability of ingredients or logistics constraints could impact the end product. AI is also used to support real-time process optimisation, helping to detect deviations earlier and stabilise production lines.

Highlighting issues before they get too big maintains product quality and availability, while simulations help the operations teams adjust plans if needed and ultimately save the business time and money. “This reduces time to market for new products, supports smoother industrialisation and helps ensure that innovation reaches consumers faster and more consistently,” adds Villain.

The system is already making a difference. Nestlé created additional packaging supply in premium pet food when demand forecasts rose sharply while production was brought forward and additional manufacturing capacity unlocked for cold coffee concentrates. In confectionery, it has removed bottlenecks and is preparing additional line capacity for new products, so supply can stay ahead of the growth curve.

But it is AI’s ability to speed up decision-making that ultimately makes it an invaluable tool for Nestlé, a business that has its arms open and ready to welcome more AI systems in the future.

“The focus is on scaling what works – responsibly – and ensuring that technology ultimately translates into better, more reliable products for consumers. The direction is clear: moving from isolated use cases to enterprise‑wide, value‑driven AI embedded in daily operations,” says Villain.