Unilever Deploys AI Digital Twins Across Global Factories, With Results in Poznan

Unilever has announced an expansion of its AI-driven digital twin program to additional manufacturing plants as part of a global partnership with Accenture. The company plans to roll out more than 40 new deployments over the next 18 months, and one plant where the technology is already delivering measurable results is the factory in Poznan, Poland, which makes Hellmann's and Knorr products.

A digital twin is a virtual model of a machine, piece of equipment, or an entire production process, continuously fed with data from real systems on the factory floor. This lets engineers monitor production lines in real time, predict failures before they happen, test different scenarios, and optimize processes without risking a shutdown of the physical line.

How it works in practice

At the plant in Raeford, North Carolina, which produces stick deodorants, the digital twin predicts as much as 95 percent of production flow constraints. That translates into a 20 percent reduction in waste and a 10 percent increase in production capacity. In India, at one of the largest cosmetics plants in South Asia, a similar system improved soap quality, cutting nonconformities by 30 percent over four years through real-time process control.

The system does more than monitor, it also learns from the data it collects over time. As workers' trust in its accuracy grows, it can gradually take over certain corrections automatically, still under human oversight. This approach differs from the full automation some companies have tried to implement in recent years with mixed results, as shown by the case of Ford, which had to bring back hundreds of engineers after a failed attempt to automate quality control.

Results at the Polish plant

Unilever's factory in Poznan, which produces mayonnaise and sauces under the Hellmann's brand as well as Knorr seasonings and stocks, is one of the plants where digital twins have already been running for some time. According to figures provided by the company, the technology has cut short production stoppages by about 20 percent and reduced the amount of waste generated by nearly 30 percent. Those are concrete, measurable numbers showing that investment in industrial artificial intelligence at Polish plants owned by global corporations is translating into real savings, not just marketing slogans.

Scaling AI across our operations is not just a technology shift, it is a commitment to creating better products, caring for sustainability, and strengthening our teams - Adam Raeburn-James, Global Vice President of Digital Business Operations, Unilever

On Accenture's side, the collaboration with Unilever is led by Nicole van Det, CEO of Accenture Netherlands and Nordics and global lead for the Unilever account. She notes that the expanded use of digital twins in manufacturing reflects Unilever's long-standing experience in supply chain management, combining technology with employee skills development.

Unilever has long been recognized as a leader in supply chain excellence, and the expanded use of digital twins in manufacturing reflects the company's continued focus on both technology and people - Nicole van Det, CEO Accenture Netherlands and Nordics

What it means for Polish industry

The Poznan deployment illustrates a pattern increasingly seen at Polish plants owned by global corporations: technology arrives first at headquarters, gets tested at select factories worldwide, and is then scaled to further locations, including those in Poland. For Polish engineers and production managers, this means having to learn to work with data-driven predictive tools rather than just traditional control systems. Companies that already have access to such solutions through their global structures gain a cost advantage over smaller, local manufacturers that do not yet have these tools.

The expansion of the digital twin program also fits into a broader trend of cautious but steady adoption of industrial artificial intelligence by large manufacturing companies. Unlike the loud announcements of full automation that in some industries have proven premature, Unilever's approach involves the system gradually taking on tasks while always leaving room for human intervention. That model may prove more durable than attempts to fully replace engineering teams with algorithms.

Over the next 18 months, Unilever plans to launch more than 40 new digital twins at factories around the world, building on existing deployments, including the one in Poznan, into a repeatable rollout template meant to be easier to scale than individually designed, one-off systems. If the results from Poznan and Raeford hold up, other Polish plants belonging to large food and chemical companies can be expected to follow a similar path.

Sources: Unilever x Accenture partnership scales digital twins (unilever.com), Unilever Scales Digital Twins Across Global Manufacturing Network with Accenture (newsroom.accenture.com), Digital twins prove themselves at the Poznan factory (epoznan.pl), Digital twins are doing their job at Unilever's Poznan factory (mamstartup.pl)