The Physical AI Revolution: Europe's Strategic Push in the Global Humanoid Robotics Race

Session: The Innovation and Future of Humanoid Robotics through Industry-Academia Collaboration

Speakers:

• Karsten Lemmer, Executive Board Member for Innovation, Technology Transfer and Research Infrastructure, German Aerospace Center (DLR)

• Alin Albu-Schäffer, Director of the Institute of Robotics and Mechatronics, German Aerospace Center (DLR)

• Dr. Kal Mos, Head of Research and Predevelopment, Siemens AG

Session Summary: Driven by severe European demographic labor shortages and advancements in Vision-Language-Action models, humanoid robotics is rapidly transitioning from a media hype into an industrial necessity. Scaling these complex machines for reliable factory deployment requires overcoming immense physical data bottlenecks and integrating them seamlessly into existing automation pipelines. Securing Europe's competitive edge ultimately demands strict collaboration between foundational research institutions like DLR and industrial scaling partners like Siemens AG, supported by robust political funding.

Artificial intelligence is officially transitioning from the digital realm into the physical world, taking the highly tangible form of humanoid robots. While viral videos of robots—such as Tesla's Optimus performing somersaults or a machine setting a half-marathon world record—routinely capture public attention, industry leaders emphasize that dancing and sprinting do not create true industrial value. At a recent high-level panel featuring experts from the German Aerospace Center (DLR) and Siemens AG, the focus shifted sharply from media novelty to strict industrial necessity.

Karsten Lemmer, Executive Board Member for Innovation, Technology Transfer and Research Infrastructure at the German Aerospace Center (DLR), noted that media visibility for humanoid robotics surged by a staggering 715% in the year 2025. However, this technological push is not merely a hype cycle. Data from DLR's foresight department and the data platform MapG reveals a distinct and significant upswing in global patent applications since 2023, indicating that the Technology Readiness Level (TRL) of humanoids is steadily increasing toward real-world viability.

The Demographic Imperative and the "ChatGPT Moment"

Alin Albu-Schäffer, Director of the Institute of Robotics and Mechatronics at the German Aerospace Center (DLR), highlighted the massive societal driver accelerating this technology: a severe demographic shift across Europe. Over the next five years, Europe will lose 10 million workers to retirement, a number that is projected to double to 20 million over the next decade. Because approximately half of these retiring individuals perform manual labor, this demographic cliff creates an immediate market demand for millions of robotic workers.

To meet this demand, the robotics industry is actively pursuing its own "ChatGPT moment". Powered by Vision-Language-Action (VLA) models, the expectation is that users will soon be able to issue direct verbal commands to a humanoid robot—such as instructing it to assemble IKEA furniture or change a car's air filter—and the robot will autonomously execute the task.

Global Competition and Market Positioning

The race to capture this lucrative market is heavily contested by global superpowers. Karsten Lemmer provided a detailed breakdown of the global landscape:

• Asia-Pacific: Currently dominates the market with an approximate 40% share, largely driven by the People's Republic of China, which explicitly elevated robotics and AI to central, state-prioritized technologies in its 2026 five-year plan.

• United States: Development is primarily fueled by aggressive private sector and venture capital investment, with $2.8 billion poured into humanoid robotics in 2025 alone.

• Europe: Despite massive investments abroad, Europe remains a formidable contender. Out of the top 500 organizations in the humanoid robotics sector, the US hosts 118, China hosts 67, and Europe hosts over 100. Germany alone is home to 42 top entities, including industry giants like Siemens AG and the German Aerospace Center (DLR).

• 
  

The Roadmap to Commercial Integration

Alin Albu-Schäffer noted that while previous generations of collaborative robots (cobots) took about 15 years to develop from prototypes into market-ready products, humanoid robotics is currently only in the "first 5 kilometers of a marathon".

To illustrate the path forward, Karsten Lemmer shared DLR's projected timeline for the real-world deployment of these highly adaptable machines:

• 5+ Years: Strong demand and commercial availability are expected in the industrial, production, and logistics sectors.

• 5 to 15 Years: Deployments will expand into service industries, retail, and the healthcare and nursing sectors.

• 10 to 20 Years: The technology is expected to reach the household and smart home sector at economically viable consumer prices.

• Long-term Applications: Integration into rescue operations, defense, and space missions.

  
  

Critical Industrial and Technological Challenges

Dr. Kal Mos, Head of Research and Predevelopment at Siemens AG, noted that engineering a functional industrial humanoid is significantly more complex than building an electric car. To physically embed intelligence into humanoid forms like DLR's TORO, the industry must overcome four core challenges:

• Continuous Model Improvement: Foundational AI models must be continuously enhanced, as they are the direct source of the robot's intelligence, allowing it to perceive, reason, and act flexibly.

• Absolute Reliability: While an accuracy rate of 70% to 90% is excellent for a laboratory demonstration, deploying a Robotic Foundational Model (RFM) on a live production floor requires absolute reliability in the "high 90s".

• Seamless Integration: Individual robots will not solve factory-wide problems unless they are flawlessly integrated into entire, pre-existing pipelines of industrial automation.

• The Data Bottleneck: Unlike Large Language Models (LLMs) that pull from massive amounts of pre-existing internet text, there is no "internet of motion". Physical data is incredibly difficult to acquire and must be generated through manual teleoperation or sophisticated synthetic data generation in virtual environments.

The Necessity of Ecosystem Collaboration and Political Will

To transition technologies out of the laboratory and into the real world, strict collaboration between academic research and private industry is mandatory. Research institutions generate fundamental breakthroughs, but they require industrial scaling partners like Siemens AG to validate these concepts, provide real-world testing environments, and transform them into marketable, highly reliable products.

Furthermore, robust political will and funding are absolute necessities. Leveraging frameworks like Germany's "High-Tech Agenda" and establishing Europe-wide networking for AI-based robotics—including planned lobbying efforts at the European Parliament coordinated by Alin Albu-Schäffer—will be crucial in ensuring that Europe remains a leader. As the technology advances, humanoids will need to prove they are safe, highly adaptable across varied terrains like asphalt, laminate, and sand, and entirely reliable in environments originally built exclusively for humans.