The race for artificial intelligence is no longer played out only in data centers. After three years dominated by language models, computational infrastructures and software agents, a new frontier is now attracting capital, that of physical intelligence.
The Barcelona startup THEKER announces a fundraising of 85 million dollars, or approximately 73 million euros. The operation constitutes the largest Series A ever carried out in European robotics. Co-led by CRV, the tour also brings together Samsung, Cathay Innovation, LVMH, 20VC, Henkel, Korelya and Sonae.
Taken in isolation, this financing might appear as a new fundraising in an already announcement-rich sector, in a broader perspective it illustrates a much deeper transformation: the emergence of a new generation of European companies seeking to apply advances in artificial intelligence to real-world operations.
Just a few hours after the announcements of NEURA Robotics in Germany, the rise of autonomous defense players like Alta Ares or the industrial demonstrations of Chinese players led by Unitree Robotics, THEKER’s operation confirms that robotics is becoming one of the main areas of application of AI.
Artificial intelligence leaves screens
The first phase of the AI revolution was to automate cognitive work. Models developed by OpenAI, Anthropic, Google DeepMind or Mistral AI have demonstrated their ability to produce text, generate code, analyze documents or assist in complex decisions. At the same time, hundreds of billions of dollars have been committed to building computing and energy production infrastructure to support this growth.
This first phase made it possible to create an intelligence capable of understanding. The second seeks to create an intelligence capable of action.
This is precisely the ambition of general robotics. Where traditional industrial robots perform defined tasks in perfectly controlled environments, new generations of systems seek to adapt to real-world variations. They must recognize different objects, manage unexpected situations, modify their behaviors and learn from their experiences.
In other words, the same principles that allowed the emergence of foundational models are now applied to robotics.
Building a physical intelligence platform
Founded in 2022 by Carla Gómez Cano and Jiaqiang Ye Zhu, THEKER develops robots designed to operate in complex industrial environments without requiring permanent reprogramming.
The company claims an “AI-native” approach in which adaptation capabilities constitute the heart of the system. The robots are capable of adjusting to production changes, manipulating different references or integrating new operational parameters without going through the integration cycles traditionally associated with industrial robotics.
But THEKER’s real product is probably not the robot itself. As in generative artificial intelligence, the value tends to shift to the software layer. Visual perception, decision-making, motor control, data collection, continuous learning and orchestration of operations form a platform that more closely resembles an operating system of the physical world than a simple industrial machine.
The comparison with language models appears more and more relevant. Where OpenAI aims to become a universal cognitive layer, companies like THEKER seek to become a universal operational layer capable of interacting with industrial processes.
A fundraising that goes far beyond the case of a Spanish startup
The composition of the round table is probably the most revealing information.
CRV is historically associated with several major American software successes, including DoorDash, Mercury and Vercel. His arrival in the capital of THEKER reflects the growing interest of software investors in robotics.
For a long time, hardware was considered difficult, capital-intensive and less scalable than software. AI is gradually changing this equation. Investors are now seeing platforms that can drive network effects driven by data and continuous learning.
Samsung’s presence also sends an important signal. The Korean group is making its first investment here in a Spanish startup. This choice reflects the growing attention paid to the convergence between electronics, artificial intelligence and automation.
But the most surprising investor undoubtedly remains LVMH. At first glance, the merger between a luxury group and an industrial robotics startup may seem unexpected. However, it reflects a profound evolution in industrial priorities. Management of logistics flows, quality control, product customization, flexible automation of workshops or optimization of operations are all subjects likely to sustainably transform industrial models in the luxury sector.
LVMH’s investment illustrates a conviction that goes well beyond its activity: robotics could become a strategic infrastructure comparable to what artificial intelligence represents today for digital functions.
Europe seeks to move up the value chain
For several decades, Europe has held a strong position in industrial automation. Germany has built world leaders in industrial equipment. Switzerland has established itself in several specialized segments. France has developed recognized know-how in industrial software and certain robotic technologies. But integrated robotic platforms capable of competing with American or Chinese ambitions remained rare.
This situation is evolving rapidly; several European players are now emerging around THEKER who are pursuing comparable ambitions. NEURA Robotics is developing a complete robotics ecosystem built around its Neuraverse platform. ANYbotics continues its expansion into autonomous industrial inspection. Wandercraft is expanding its work on advanced robotic systems while Exotec continues to demonstrate European capacity to industrialize large-scale robotic platforms.
Against them, American references remain considerable. Figure AI, Physical Intelligence or Skild AI attract several hundred million dollars to build future foundational models of robotics.
The difference is that Europe no longer only seeks to produce components. It is now trying to build complete platforms.
The real strategic asset is no longer the robot
As in generative AI, the debate often focuses on visible models or machines.
However, the value could be concentrated elsewhere, the leaders in artificial intelligence now have a considerable advantage because they accumulate massive volumes of data and have the infrastructure necessary to exploit them, robotics follows a similar logic.
Each movement made in a warehouse, each manipulation carried out on a production line, each error corrected and each environment explored produces new data. This information helps improve system performance and gradually reduce deployment costs.
This dynamic creates a learning loop comparable to that observed in language models. The central question then becomes less that of the robot than that of the data accumulated by this robot.
Future industry leaders could be those who control the largest volumes of real-world physical interactions.
General robotics still needs to prove itself
However, investors’ enthusiasm should not mask certain questions. The term general robot has become omnipresent in the presentations of startups in the sector. However, industrial history has often demonstrated the economic superiority of perfectly optimized specialized systems.
A production line does not need a robot capable of performing a hundred different tasks. It needs a robot capable of perfectly executing a critical operation millions of times without error.
The industry imposes constraints that the software world knows little about: security, availability, maintenance, precision and reliability.
The next challenge: deployment
The other blind spot in the sector concerns industrialization, technological demonstrations are increasing. Fundraising too.
But the essential question remains largely open: how many robots are actually deployed on a large scale? The recent history of artificial intelligence shows that technology is not always the main obstacle. In many cases, the challenge lies in integrating with existing systems, user adoption and business process transformation. Robotics could follow the same trajectory.
The challenge of the coming years will not only be to build more intelligent robots. It will consist of deploying them in thousands of factories, warehouses and infrastructures where each interruption of service has an immediate economic cost.
Faced with the United States, China is accelerating
Another question remains relatively absent from European debates. Most analyzes focus on American competition, yet Chinese progress could prove just as decisive.
Players like Unitree Robotics or UBTECH Robotics benefit from unique proximity to the country’s supply chains, manufacturing capacities and industrial infrastructure.
China already has a dominant position in several strategic hardware segments. If it can replicate this advantage in AI-powered robotics, the competition could quickly move beyond just technology.
