Royal Swinkels

When intelligence takes shape: the rise of physical AI


Robots that learn, adapt, and collaborate: physical AI is transforming industries and work. Europe must act now to lead.


In brief

  • Physical AI enables flexible, self-learning robots that are reshaping industries and jobs.
  • Europe’s legacy data offers a unique advantage, but rapid action is needed to stay ahead.
  • Creativity, empathy, and innovation will be the most valuable human skills as AI takes over repetitive tasks.

Imagine robots that don’t just follow commands, they learn, adapt, and collaborate. This isn’t science fiction anymore. It’s the dawn of physical AI, and Europe has a small window to lead or risk falling behind. But leadership requires more than ambition, it demands awareness of the risks and adoption barriers that come with this disruptive shift: safety, compliance and liability, cybersecurity for connected robots, and integration with legacy automation. Dr. Adrian Reisch delivered a powerful message during EY AI Week: “Proven systems are no longer fit for purpose. Physical AI is disruptive, and everyone will be impacted.” Here’s why this matters and what you need to do now.

What is physical AI?

The term is often used conceptually, but executives need clarity: What sets physical AI apart from traditional robotics and AI, and what are its core capabilities in perception, dexterity, autonomy? In essence, physical AI combines robotics, sensing, simulation, and self-learning control systems that operate in the physical world without constant human programming.

Why physical AI is the answer

For decades, industrial automation relied on rigid, pre-programmed systems. They worked well for high-volume, low-mix production. But today’s reality demands agility: shorter product cycles, customized orders, and rapid reconfiguration. Static robots can’t keep up.

Physical AI changes the game. Robots can now program themselves, handle flexible parts, and adapt to new tasks without weeks of re-engineering. This means:

  • High mix, low volume production becomes viable.
  • Setup times shrink dramatically.
  • Robots move beyond assembly lines into service, maintenance, and logistics.

Europe faces demographic pressure: millions of skilled workers retiring in the next decade. Physical AI isn’t optional; it’s a strategic necessity. Dr. Adrian Reisch: “AI is transformative and will impact everyone’s life. The United States is currently ahead, especially when it comes to hyperscalers and large language models, what you might call GenAI. They have a significant head start and far more capital than we do. However, Europe has a unique advantage: legacy data. We have decades of experience in high quality engineering, manufacturing, and using products. By training AI on this legacy data, we can build powerful industrial foundation models. This is our opportunity and our unique selling point.

But time is of the essence. China is catching up rapidly. While they may not have the same legacy, they are investing heavily in designing, manufacturing, and deploying new products, generating data at a much faster pace than we are. Europe’s opportunity lies in leveraging this legacy data to build industrial AI-models. At the same time, we’re facing a demographic shift: baby boomers are retiring, and we must do more with fewer people. For European companies to regain competitiveness, physical AI is inevitable: it’s about bringing products to market faster, achieving more with less, and capitalizing on trends like autonomous driving, software-defined manufacturing, and humanoid robots.”

Our pursuit of perfection can actually hinder progress.

The hockey stick moment

Look at the numbers. Patent filings in AI, robotics, IoT, and extended reality have exploded. GenAI accuracy scores jumped from 3.3 to over 40 in just 18 months. Humanoid robots? Analysts* predict 60% annual growth, reaching 650 million units by 2050. Serial production of humanoid robots is expected within two to four years. Costs will drop. Adoption will accelerate. If you wait, you’ll be playing catch-up while competitors lock in talent, platforms, and market share.

 

Dr. Reisch explains what’s holding companies back. “In Europe, especially Germany, we place a high value on reliability and quality, often perfecting solutions before implementation. While this careful approach has its merits, it can slow us down. To stay competitive, we need to embrace agility: build an 80% solution with 20% of the effort, then iterate and improve. Accepting some imperfection is essential, companies like Tesla succeed by launching early and refining as they go. Our pursuit of perfection can actually hinder progress.”

 

Robotics-as-a-service

Dr. Reisch: “I recommend embracing EY’s motto: shape the future with confidence. If you don’t take charge of your own future, someone else will. United, Europe is the world’s largest economy, and Germany still ranks third globally. We have all the resources we need; we just need to act.” A compelling example comes from a service company maintaining industrial plants. Traditionally, their revenue grows linearly, more contracts require more employees, with margins around 10–15%. With robotics-as-a-service, once the infrastructure is in place, one trained robot can instantly scale to thousands. Margins multiply by four or five, and growth becomes exponential. This isn’t just theory: it’s happening right now.”

 

Why robotics-as-a-service will dominate

Cash is king. Capex restrictions are real. Subscription models win: just ask Netflix or SaaS providers.

 

Robotics-as-a-service offers:

  • Lower upfront costs
  • Scalable deployment
  • Continuous updates and shared learning across fleets

Combine this with Europe’s manufacturing strengths—precision, reliability, system integration—and the opportunity is massive. But only if we act fast.

 

Disruption across industries

In the next five years, physical AI will become an integral part of daily life and work. We’ll see the emergence of cognitive robots: humanoids capable of adapting, learning, and training themselves. These robots will take on service tasks that are increasingly difficult to fill with human labor, such as assembling components in factories, cleaning offices, refilling coffee machines, and managing waste. Dr. Adrian Reisch: “In sectors like healthcare, humanoid robots will support facility management by sorting laundry and maintaining cleanliness, freeing up staff for more specialized work. While these developments may not be mainstream by 2026 or 2027, they are expected to become reality before 2030.”

 

Physical AI’s impact will extend far beyond manufacturing. More than half of all use cases will span industries such as logistics, healthcare, energy, and retail. Imagine robot dogs inspecting fire extinguishers, drones monitoring pipelines, and humanoids assisting in care facilities. The potential market is vast, and the transformation will be visible across the entire economy.

Preparation starts at the top: set a bold vision and launch lighthouses: small, value driven pilots, because transformation runs at two speeds.

The human factor

Culture beats fear, technology alone won’t win. Culture will. Dr. Reisch drew inspiration from Formula 1 and McLaren CEO Zak Brown: “Talent is what’s most important. While we’re in the technology business, it’s our people that develop and create the technology.” High performance requires trust, collaboration, and fun, not fear or blame. Innovation thrives when teams feel safe to experiment. Physical AI is new territory. There’s no playbook. Leaders must create environments where curiosity beats caution.

The greatest challenge is shifting the collective mindset: embracing a culture that allows for failure, encourages experimentation, and values learning over perfection. Business can and should be both enjoyable and successful. Rapid progress comes from allowing failure, learning quickly, and iterating faster. Success in physical AI also depends on combining deep domain knowledge in manufacturing and engineering with AI expertise. Dr. Reisch: “You can’t simply turn an engineer into an AI expert, or vice versa; true progress requires multidisciplinary teams working together, even if they speak different professional languages.

Preparing organizations for this future starts at the top. Leaders must set the example by accepting that nobody is perfect, except machines at their specific tasks. Companies need to operate at two speeds: maintaining a clear long-term vision while simultaneously launching small, value-driven pilot projects, lighthouses, and building on minimum viable products.”

Speedboats and strategy

The revolution demands two speeds:

  1. Strategic target picture: Define your long-term vision. What capabilities will you need to scale?
  2. Speedboats: Launch quick wins that deliver tangible value. Train robot dogs before humanoids. Use GenAI for predictive maintenance. Build momentum.

Think of it as planning a journey: set your destination, map your route, choose your transport, and bring the right crew. EY’s AI Lab helps clients do exactly that, showing the art of the possible and designing roadmaps that balance ambition with action.

Creativity, empathy, and innovation

The window is closing: are you ready? Physical AI will reshape industries, jobs, and businessmodels. It’s not about replacing humans; it’s about freeing us from repetitive tasks. Ultimately, the human qualities that will matter most are those machines cannot replicate: creativity, empathy, and innovation. As AI takes over routine work, people will focus on what makes us uniquely human skills that will only grow in value. But speed alone isn’t enough. Europe must lead by building the right foundations:

  • Standardize datamodels across sectors for example manufacturing and service logs
  • Incentivize cross-industry data sharing to accelerate innovation
  • Invest in foundational industrial models that enable scalable adoption

A coordinated policy and ecosystem approach will determine who defines the platforms, captures the talent, and leads the next industrial revolution.

*Source: Citigroup-studie


The EY.ai Lab

In the EY.ai Lab you can experience immersive, hands-on tailored workshops with your team that apply AI to core business processes. Guided by EY practitioners, you’ll explore real-world use cases, learn practical methods and tools, and shape solutions tailored to your needs.

EY.ai Lab promotional image


Summary

Physical AI is revolutionizing industry with self-learning robots. Europe’s legacy data offers a unique edge, but urgent action is needed. Creativity, empathy, and innovation will define future success as AI takes over routine tasks. First movers will shape the platforms and lead the next industrial revolution.


About this article

Read more

The future organization: humans at the heart of AI-transformation

Discover how leaders keep humanity central in the age of AI. Learn how to build organizations where technology empowers people. Read the full article!

How FrieslandCampina’s Data Platform paves the way for AI innovation

FrieslandCampina's data platform enables AI innovation and drives transformation from reporting to prediction and automation.

Key insights from EY AI Week: balancing innovation with a human-centered approach

Discover insights from the AI week where industry leaders explore AI's transformative impact across various sectors.