China’s dark factories: the car factory without employees

China is currently developing a new generation of car factories in which robots no longer merely assist but take charge of the entire manufacturing process. Before 2030, the country is expected to have its first completely staffless plant producing cars from start to finish - from raw sheet metal to the finished vehicle, without an employee ever touching the steering wheel.

From production line to dark factory

For decades, people and robots have worked side by side in car plants. Welding robots, paint-shop arms and logistics vehicles have long automated many tasks. Final assembly, however - the precise fitting together of every component - has until now remained largely the preserve of people.

That final area of human dominance on the factory floor is now beginning to disappear. Manufacturers in China are planning their first plants with no conventional jobs at all. The lines will keep moving, yet the production halls will be almost empty.

“Robots weld, paint, screw and inspect - without breaks, fatigue or holidays.”

Internationally, the idea has acquired a fitting name: so-called “dark factories”. These are production sites that could, in principle, operate without lighting because nobody needs to see what is taking place inside. The machinery navigates using sensors, cameras, laser scanners and central control computers supported by AI.

Why China is accelerating dark factories

Over recent years, China has invested heavily in robotics and artificial intelligence. Service robots in public spaces, autonomous patrol vehicles and intelligent traffic management are no longer visions of the future there; they are already part of everyday life in many cities.

This expertise is now moving into heavy industry. New car factories are being designed from the outset to require barely any staff. Sensors monitor every stage, while AI systems intervene if a component does not fit precisely or a machine deviates from its target setting.

• 24/7 operation without shift changes or night-shift premiums
• consistent quality because robots work with precision and repeatability
• flexible conversion of production lines through software updates
• substantial savings on wages, social contributions and building costs

Industry experts expect China’s first fully automated car factory to enter series production before the end of the decade. The US is regarded as a close challenger, while Europe is somewhat behind but does not intend to remain so permanently.

Automation as a response to cost pressure

The global automotive industry is in the middle of the most expensive transformation in its history. Electric powertrains, new software platforms, connectivity and autonomous driving all consume billions. At the same time, margins are falling in many markets as competition becomes increasingly fierce.

This is exactly where fully automated plants come in. Their purpose is to reduce the cost per vehicle and significantly shorten development times. Studies such as those from consultancy firm Accenture illustrate the potential behind this strategy.

“Advanced automation can reduce development and time-to-market periods by up to 50 per cent.”

Companies that can move a car from concept to sale more quickly and at lower cost gain market share. That is why not only Chinese brands are investing in this technology, but western automotive groups too.

Hyundai, Tesla & Co.: the race towards the robot factory

Hyundai is among the manufacturers pursuing automation particularly aggressively. Within just a few years, the group intends to use humanoid robots in its plants. From 2028, Boston Dynamics models are expected to move around a factory in the US state of Georgia, assisting with transport, maintenance and monitoring.

Tesla is taking a somewhat different route. The US company relies on enormous casting machines, known as Megacastings, which produce large sections of the vehicle frame as a single piece. The fewer individual parts involved, the easier it is to automate everything else. At the same time, Tesla is developing the humanoid robot Optimus, which is also intended for eventual use in manufacturing.

German manufacturers are not standing still either. BMW, Mercedes-Benz and other premium brands are already testing highly automated lines where only a small number of skilled employees are directly involved in production. They step in only when something goes wrong or particularly delicate work is required.

How processes are changing in practice

A look at standard processes shows just how extensive the change could be:

What “dark factory” means in practice

A factory with hardly any lights on may sound like science fiction. In reality, it relies on several technical building blocks:

• robot arms with tactile sensors and cameras that identify parts and position them correctly
• mobile robots that carry components between workstations
• digital twins of the factory, allowing processes to be simulated and optimised virtually
• AI systems that analyse production data in real time and automatically make adjustments when required

The “dark” aspect is more symbolic than mandatory. In practice, many of these plants will at least be dimly lit, if only to allow maintenance work. The central idea nevertheless remains the same: people are no longer a permanent part of the process and, ideally, only attend for maintenance, supervision and further development.

Opportunities and risks for jobs

For manufacturers, fully automated plants may sound ideal. From employees’ perspective, the situation looks very different. If entire production sections can function without staff, traditional industrial jobs face enormous pressure.

At the same time, new roles are emerging. Data analysts, robotics engineers, maintenance specialists and IT security experts will be in greater demand. Required skills are shifting away from shift-based factory work towards highly trained technicians and software professionals.

For high-wage countries such as Germany, extensive automation could paradoxically become an advantage as a manufacturing location. When labour accounts are substantially reduced, the wage gap with low-cost countries becomes less significant. Companies may be more able to keep production in their home market instead of moving it entirely to Asia - provided the necessary technology is available.

What car drivers and customers can expect

At first glance, car buyers may notice little difference between a vehicle made in a conventional factory and one produced in a dark factory. In practice, however, several consequences could reach the customer directly:

• shorter model cycles because new variants can be introduced more quickly
• more consistent quality, as faults are detected earlier and more systematically
• potentially lower prices if manufacturers pass on cost savings
• greater standardisation of platforms and components

It will be interesting to see how openly manufacturers deal with the issue. Some may present glossy, staffless plants as high-tech showpieces. Others may be more restrained in order not to intensify debates about job losses.

Terms worth knowing

Anyone following this development will repeatedly encounter certain key terms. Two appear particularly often:

Digital twins: These are virtual representations of a factory or a product. Engineers can use the model to test processes before they are introduced in the real world. For example, it can be used to assess how a new robot cell affects the entire flow of materials.

Predictive Maintenance: This means proactive maintenance. Sensors provide early warnings when a component is wearing out. The AI recommends an appropriate time for replacement so that unplanned stoppages can be avoided. This plays a central role in a plant with no staff on the factory floor.

It has not yet been conclusively established whether China will truly become the first country to launch a fully functioning car factory without employees. What is clear, however, is that the direction has been set, and leading manufacturers around the world are aligning their strategies accordingly. Those who wait too long risk doing no more than following behind in the next technological leap.