Inside the virtual factory: where companies simulate workers before they exist
Inside a BMW Group Virtual Factory (photo: BMW Group)
Before a new production line is built, before workers are trained, and in some cases before a factory physically exists, companies are testing how people will move, interact and perform inside fully virtual environments. This approach has been developing for several years, but recent deployments show it moving beyond pilot projects into scaled industrial use. From automotive and electronics to logistics and industrial engineering, firms are combining digital twins—virtual replicas of physical systems—with artificial intelligence (AI) and simulation tools to model entire workplaces, including human workers, before real-world implementation.
From virtual models to operational systems
At BMW, the “Virtual Factory” has been rolled out across its global production network. The company uses digital twins to simulate production processes, including manual work steps, allowing planners to test how workers interact with tools and assembly lines before production begins.
The approach is already being applied in new facilities such as the Debrecen plant in Hungary, where production planning has been extensively tested in a digital environment prior to ramp-up. According to the company, this enables earlier identification of design issues and reduces the need for physical rework during installation.
Other industrial technology providers are embedding similar capabilities into their platforms. Siemens integrates human workflow simulation into its manufacturing software, while Dassault Systèmes offers “virtual human” modelling within its 3DEXPERIENCE platform, widely used in aerospace and automotive design.
The shift is not confined to a single company. Simulating workers in digital factories is becoming part of standard industrial planning.
Designing workflows before they exist
The operational impact lies in how workflows are created.
Traditionally, production lines are refined after launch. Simulation shifts that process forward. Engineers can test layouts, task sequences and timing before equipment is installed or staff are trained.
At Ford Motor Company, digital human modelling is used to assess ergonomics in assembly processes, allowing engineers to reduce strain and optimise workflows before implementation. Toyota applies similar principles as part of its production system, using simulation to refine human-machine interaction.
These systems allow companies to identify bottlenecks, test alternatives and standardise processes across sites without interrupting operations.
Simulating workers, not just machines
Digital twins have long been used to model machines and logistics systems. The inclusion of workers marks a shift in how these tools are applied.
At BMW’s Regensburg plant, teams use virtual reality to explore assembly processes years before production begins. Dassault Systèmes’ modelling tools allow companies to simulate reach, posture and movement within those environments.
The focus is practical: identifying inefficiencies, improving ergonomics and reducing workplace risk. Early-stage simulation can inform workstation design, tool placement and task sequencing before they are fixed in the physical world.
These models remain approximations. They rely on predefined parameters and datasets and cannot fully capture the variability of real-world human behaviour.
A broader shift across industry
The use of simulated workers is expanding beyond automotive manufacturing.
In logistics, Amazon uses simulation models of fulfilment centres to test how human workers and robots interact within shared environments. These models help evaluate layout changes and operational strategies before implementation.
In industrial analytics, SAS Institute has demonstrated simulation environments that incorporate workers, vehicles and automated systems, allowing companies to analyse safety scenarios and operational performance.
Across sectors, the direction is consistent: human activity is being modelled as part of the system itself.
Where simulation meets automation
Simulation is closely linked to automation strategies.
BMW has begun piloting humanoid robots in production environments, including at its Leipzig plant, focusing on repetitive or physically demanding tasks. Similar dynamics are visible in logistics, where Amazon continues to expand robotic systems within warehouses.
In both cases, simulation plays a role in testing how humans and machines will operate together. Companies model movement patterns, task allocation and safety protocols before introducing automation into live environments.
Training and behavioural modelling
Simulation is also being applied to training.
Virtual environments allow workers to familiarise themselves with layouts and processes before entering live production settings. This can reduce onboarding time and improve consistency across sites.
Companies such as Virti develop virtual training environments where employees interact with simulated scenarios. While these systems are primarily used for training, they contribute to a broader shift towards modelling human behaviour as part of operational planning.
Limits, data and ethical considerations
Despite advances, simulating human behaviour remains a technical and practical challenge.
Unlike machines, human performance varies based on experience, fatigue and context. Most industrial applications remain focused on specific use cases—such as ergonomics or workflow optimisation—rather than full behavioural replication.
There are also data and governance considerations. Modelling workers requires datasets that capture movement and performance, raising questions about how this data is collected, anonymised and used.
Representation is another constraint. Models built on limited datasets may not reflect the diversity of real workforces, which can influence design outcomes.
Simulation also tends to prioritise measurable efficiency. Less visible aspects of work—such as informal collaboration or adaptability—are harder to capture.
For these reasons, companies continue to combine digital simulation with physical testing and worker input.
A shift in industrial decision-making
What distinguishes current developments is the role of simulation in decision-making.
Digital models are increasingly used to test scenarios before implementation, influencing decisions about factory layout, workforce planning and operations. This allows companies to explore alternatives without disrupting production.
Simulation reduces uncertainty, but it does not eliminate it. Its effectiveness depends on data quality, assumptions and interpretation.
Meanwhile, the broader shift is conceptual.
Workers are no longer treated as external users of industrial systems but as components within them. Their movements and interactions are modelled alongside machines and processes.
This enables more precise planning, while also introducing new forms of measurement into the workplace.
The direction of travel
The use of simulated workers in virtual factories is now established across multiple industries. Companies including BMW, Siemens, Dassault Systèmes, Ford and Amazon are applying these tools in production planning, training and automation.
Factories are now designed in code before they are built in steel—and workers are part of that code. As simulation moves deeper into decision-making, it begins to define not just how work is done, but what work looks like in the first place. The systems being built today will not only reflect reality; they will increasingly set its parameters.
Liked this article? You can support our independent journalism via our page on Buy Me a Coffee. It helps keep MoveTheNeedle.news focused on depth, not clicks.