BMW embraces humanoid robots to revolutionise European car manufacturing

BMW is poised to be the first car manufacturer in Europe to implement humanoid robots on its production line, marking a major change in automotive manufacturing. Two robots, named Aeon and developed by Hexagon Robotics, will start operations at the company’s Leipzig factory from summer this year, after successful test deployments. Standing 1.65 metres tall and weighing just 60 kilograms, the robots have been designed to work alongside human employees in existing factory environments without requiring costly restructuring of assembly lines. The move represents a fundamental change in how the industry approaches automation, with BMW executives describing humanoid robots as “the future of automotive production” due to their ability to function within workspaces designed for humans.

The change from conventional automated systems to anthropomorphic machines

The automotive industry has utilised robotic automation for many years, with fixed arm installations and specialised machinery becoming the cornerstone of modern factories. However, this traditional approach has always required manufacturers to redesign their entire production lines around the robots themselves—a expensive and labour-intensive process. As Bill Ray, distinguished VP analyst at Gartner, notes, when industrial robots command million-pound price tags, companies had little option except to reorganise their factories to house them. This core limitation influenced manufacturing strategy for generations, trapping factories into fixed, unchanging systems that were expensive to alter or upgrade.

The economics of robotics have shifted dramatically in recent years. As the cost of humanoid robots has fallen significantly, the calculus of factory investment has changed completely. Rather than redesigning costly production lines to fit robots, manufacturers can now install robots that fit smoothly within existing human-focused environments. This represents a paradigm shift in industrial thinking—robots must now adjust to factory environments, not the other way around. Michael Nikolaides, head of process management and digitalisation at BMW, emphasises this shift: “When a robot costs 17 million, you’d restructure your factory around the robot, but it doesn’t anymore.” This flexibility makes humanoid robots far more attractive for European firms facing tight margins and current infrastructure limitations.

• Humanoid robots fit into existing workspaces designed for human workers
• No expensive factory redesign required for deployment
• Reduced unit expenses make adaptation strategy cost-effective
• Same size and capabilities as human workers allow for straightforward implementation

How Aeon robots acquire knowledge and adjust on the factory floor

The Hexagon Aeon robots operating within BMW’s Leipzig factory are not just equipped with rigid instructions—they develop and evolve through advanced AI methods that mirror how human workers acquire expertise. Standing 1.65 metres tall and weighing just 60 kilograms, these humanoid machines are outfitted with 21 cutting-edge sensors including cameras, radar, microphones, and force and torque sensors that permit them to perceive their environment with remarkable precision. This sensory richness enables the robots to grasp complicated assembly operations and adapt appropriately to variations in their work, making them considerably more flexible than traditional industrial automation.

The training process itself represents a breakthrough in how machines learn manufacturing skills. Rather than depending exclusively on pre-programmed movements, BMW’s engineers implemented a hybrid approach combining human expertise with advanced simulation technology. Nvidia’s digital twin software allowed engineers to develop an exact virtual replica of the Leipzig factory, where robots could practise tasks multiple times in a safe environment. This integration of practical human expertise and virtual experimentation accelerates the learning process dramatically, allowing robots to master complex assembly tasks in a fraction of the time traditional programming would require.

Training methods combining human knowledge with machine learning

Teleoperation represented a crucial part of the Aeon robots’ training programme at BMW. Engineers equipped human workers with motion-capture sensors, then applied this data to teach robots how to execute delicate tasks such as grasping parts. By recording the natural variability in how humans work with objects—the fine-tuning, the multiple grip approaches, the force applications—the robots learned not just a single method to execute a task, but the full spectrum of approaches that human workers typically use. This human-led learning method ensures the robots can navigate genuine manufacturing situations with authentic versatility and flexibility.

Reinforcement learning complemented the teleoperation training, allowing robots to refine their techniques through iterative practice. Within the digital twin environment, the Aeon robots were given production assignments and executed thousands of virtual trials, identifying the most efficient and effective solutions. Arnaud Robert, head of robotics at Hexagon, emphasises imitation learning as especially significant—where robots learn by observing how work is completed, either through multi-angle video recordings or sensor data from human workers. This method has the potential to reduce training time from months to mere days, fundamentally accelerating deployment timescales.

• Teleoperation records human physical gestures for realistic task learning
• Digital twin modelling enables thousands of practice iterations in a controlled manner
• Imitation learning reduces training time from several months to just days

Broad-based uptake points to change coming

BMW’s groundbreaking partnership with Hexagon’s Aeon robots represents a significant turning point for European manufacturing. As labour costs rise and skilled workers become harder to find across the continent, automotive manufacturers encounter mounting pressure to embrace automation solutions that integrate seamlessly with existing production environments. The humanoid approach tackles a pressing industry challenge: traditional fixed automation requires expensive factory redesign, whereas human-shaped robots fit into current workflows without demanding substantial capital restructuring. This economic advantage has attracted the attention of competitors, signalling that humanoid robots may soon become commonplace across major European car plants, fundamentally transforming how vehicles are assembled.

Tackling staff concerns and practical expectations

The arrival of humanoid robots into the European automotive sector inevitably generates discussion about job security and workforce displacement. BMW and sector specialists have been deliberate in positioning the Aeon deployment not as a replacement for human workers, but rather as a solution to fill labour gaps and undertake routine, physically taxing work. Michael Nikolaides highlights that the robots are engineered to operate with human workers, boosting output rather than removing positions. The automotive sector, already contending with labour skill gaps and an older workforce, considers humanoid robots as a supplement to human work rather than a full-scale substitution plan.

However, grounded assumptions are crucial. Arnaud Robert acknowledges that whilst the ultimate vision involves machines acquiring capabilities by simply observing human workers, this prospect is “probably a year or two out.” Current deployment requires significant training and refinement period, and robots still cannot match the adaptability and problem-solving capabilities of skilled personnel. BMW’s staged implementation approach—beginning with a pair of robots at Leipzig—suggests the company recognises the requirement for measured introduction and sustained development. This deliberate speed allows manufacturers to measure actual consequences on labour trends whilst enhancing capabilities before broader deployment across European factories.

The value of robot nomenclature and shaping how people view them

Naming the humanoid robot “Aeon” as opposed to a sterile designation reflects a wider sector approach to humanise automation and shape how the public views them. By giving robots identifiable names, manufacturers establish a story of partnership instead of substitution, subtly shifting how workers and the public perceive these machines. This approach acknowledges that acceptance of industrial automation depends partly on mental framing—workers are more inclined to accept a “Aeon” as a colleague than an sterile “Unit 7”. As humanoid robots become commonplace in factories across Europe, this deliberate consideration of naming and interaction will grow more vital for sustaining broader acceptance.

• Named robots promote understanding of teamwork and cooperative working
• Human-centred naming helps workers accept automation with greater acceptance
• Public messaging shapes sustained acceptance of factory automation