Bridging AI & Physical Processes: How Unilever Embraced Digital Twins

Creating digital twins opens up a whole new world of use cases for AI agents, and gives them the levers to pull to improve operational efficiency.

Digital twins - virtual models of a complex system - can be very powerful tools for managing, maintaining and improving anything from a physical product to a complicated logistics network or supply chain. Prevalent in manufacturing and industry for the past decade or more, digital twins are becoming easier and more cost effective to create, which means they are increasingly being applied to business processes as well as large, expensive objects and machines.

Bridging virtual and physical worlds

Digital twins will likely become an important bridge between AI and the real, physical world of business as agent systems develop. Building a ‘world model’ is a huge challenge for AI tools today in areas such as autonomous driving, image analysis and logistics. Without a model of the world, AI can struggle to interpret and understand visual and operational data in a way that people take for granted. This is one of the biggest limitations of the current era of LLMs, and a key factor in their hallucinations and blind spots when it comes to tasks involving puzzles, riddles and maths.

A digital twin is a model so specific to its context that it can serve as a local world model for many purposes. Once built and running, there is huge scope to apply AI to monitoring, analysis and even action in response to changes in operating conditions.

But like so many other areas of AI readiness, the organisational and operational challenges in creating such a world model from within existing process-driven, bureaucratic firms are substantial. So it is interesting to look at the experiences of some early adopters to learn lessons on how to create digital twins of complex processes and systems as a step towards addressing the many AI use cases that exist in this area.

Unilever’s digital twin story

Unilever, like other industry leaders such as Siemens, General Electric, and Procter & Gamble, uses digital twins to improve decision-making, reduce waste, and achieve greater operational control. These virtual models that mirror their real-world counterparts provide new levels of insight across operations and help dismantle structural barriers that previously hindered digital transformation.

For Unilever, this initiative is part of its broader commitment to operational excellence, seeking to predict issues, optimise processes, and reduce resource consumption to increase efficiency and sustainability.

Among the digital twins they have already implemented are:

• Digital twin of the Dove soap production machinery, simulating plant equipment to monitor energy use and optimise production efficiency

• Digital twin of soap and detergent production processes, managing moisture levels and optimising batch processing for consistency and capacity

• Digital twin of manufacturing plant machinery, integrated with AI to reduce false alerts and improve predictive maintenance

Unilever’s Lighthouse Factory Dubai.

Adopting digital twins at scale presents challenges—from integrating complex data streams to upskilling employees in order to fully utilise these new tools. So how did Unilever manage to overcome these to create digital twins to support its production and supply chain management?

In this case study, we will explore:

• The key building blocks that enabled Unilever to successfully adopt digital twins.

• The challenges encountered in deploying digital twins at scale.

• How Unilever's culture evolved to embrace the impact of digital twins.

• What lies ahead for Unilever as it continues its digital transformation journey.

Let’s dive in.

Unilever’s history of innovation

Unilever, a global consumer goods powerhouse known for its diverse range of products—from personal care to food and beverages—remains committed to innovation and sustainability. With over 148,000 employees and a reported annual turnover of €60 billion in 2023, Unilever has consistently demonstrated operational excellence across its extensive supply chain.

Over recent years, Unilever has implemented digital twins across eight key regions, with plans to expand further, demonstrating measurable results in cost savings, energy optimisation, and increased productivity.