Germany’s VDMA engineering association is calling for Europe to build a domestic supply chain around humanoid robots as rapid growth in Chinese production turns physical AI from a research contest into an industrial scaling problem.

The association wants humanoid robotics moved higher on the German and European policy agenda, arguing that the region needs domestic production of critical components rather than relying principally on imported hardware around which European software, engineering, and integration services are assembled. Its intervention coincided with a sharp rise in shares of Chinese humanoid manufacturer Unitree following its Shanghai listing.

Global shipments of humanoid robots reached roughly 19,100 units during the first half of 2026, nearly four times the level a year earlier, according to figures cited by Reuters, with Chinese manufacturers accounting for the largest share. Those volumes remain modest beside established industrial robotics, but their growth is putting pressure on Europe’s longstanding strength in factory automation.

Europe already has substantial robotics engineering, industrial customers, component suppliers, safety expertise, and research capacity. What it has shown less clearly is an ability to convert humanoid technology into high-volume manufacturing quickly enough to compete with companies operating inside China’s much larger electronics, battery, motor, and automation supply chains.

Research strength meets a scaling problem

VDMA deputy executive director Hartmut Rauen said: “Germany and Europe must establish a comprehensive value chain featuring domestic production of critical components.” That emphasis is significant because humanoid robots are not simply AI models installed in mechanical bodies; they depend on actuators, motors, sensors, batteries, control electronics, machine vision, safety systems, and specialised manufacturing.

China already has an operational stock of around two million industrial robots, according to the International Federation of Robotics, giving domestic manufacturers a large market of factories accustomed to automation. National industrial policy has also placed AI-powered robotics among strategic technologies, reinforcing access to both supply chains and prospective customers.

European policymakers have begun putting more explicit weight on physical AI. The European Commission describes robotics as important to high-value manufacturing competitiveness and is supporting work around AI-powered robotics, interoperability, safety, and standardisation.

Public research has produced companies and specialist expertise, but commercialisation requires different capabilities from building prototypes. Production has to become repeatable, component supplies dependable, maintenance practical, and machines sufficiently reliable to operate inside working factories rather than controlled demonstrations.

Industrial customers will determine the winners

That distinction changes how humanoid systems should be assessed. Demonstrations of walking, lifting, or object manipulation can show technical progress, whereas industrial buyers purchase around throughput, reliability, safety, maintenance, integration, and the cost of replacing or supplementing existing processes.

A machine capable of performing many tasks inconsistently can be less useful than conventional automation engineered to perform one task continuously. Humanoid designs only earn their flexibility premium when that flexibility survives the operational requirements of a production environment.

Recent Techopia coverage of NEURA Robotics taking over Bosch Rexroth’s ACTIVE Shuttle platform showed one possible European route to market: combining newer physical-AI ambitions with existing products, customers, service relationships, and industrial deployments instead of building an installed base from zero.

Humanoids may have a particular advantage where environments have been designed around human bodies rather than machines. Warehouses, factories, and maintenance facilities contain stairs, shelves, tools, workstations, and doorways whose dimensions assume human movement, potentially allowing a general-purpose machine to automate work without rebuilding the entire physical environment.

However, that flexibility is commercially useful only when the machine is reliable and inexpensive enough to compete with labour or specialised automation. European opportunities therefore extend beyond producing a complete humanoid robot to motors, drives, sensors, vision systems, safety technology, digital twins, integration software, and industrial services.

The policy challenge is determining which of those layers Europe can manufacture competitively. Attempts to recreate every component domestically would be costly, while dependence on strategically important imported parts could expose robotics manufacturers to the same supply constraints already visible in semiconductors.

VDMA’s warning arrives as shipment data begins to provide harder evidence of industrial scale. If Chinese manufacturers continue widening their production and cost advantages, Europe’s robotics expertise will remain valuable, but research strength alone will not determine whose machines enter factories.