NEURA builds real-world training gyms for robots

The German robotics company NEURA has identified what it believes is the true bottleneck holding humanoid robots back from real-world usefulness. The issue isn’t a lack of intelligence, but rather a critical shortage of practical experience. To solve this, NEURA is building a network of specialized training facilities it calls “gyms,” where robots can accumulate real-world repetitions under authentic conditions.

The company recently announced a major partnership with RWTH Aachen University, one of Germany’s premier engineering institutions, to establish the NEURA Gym RWTH Aachen. This facility will span approximately 3,000 square meters at the Hightech Campus Melaten and is scheduled to open by the end of 2026. Roughly 20 university institutes will be integrated into its operations.

This is just one piece of a much larger expansion plan. NEURA has ten gyms in development, with five expected to be operational by the end of 2026 across Europe, the United States, and China. A second flagship location in Germany, the TUM RoboGym, is set to open later this year at Munich Airport in collaboration with the Technical University of Munich, backed by a €17 million investment.

The core rationale stems from a fundamental data disparity. Large language models have been trained on trillions of words scraped from the internet. Robots, by contrast, receive a minuscule fraction of that training data, and gathering it is far more difficult. While simulations are useful, they cannot perfectly replicate the friction, unpredictability, and mess of an actual factory floor.

NEURA’s solution is to have robots log real-world repetitions. The gyms combine physical practice with high-fidelity simulation, funneling the resulting training data into the Neuraverse, the company’s cloud platform that connects robots, developers, and clients. Founder and CEO David Reger put it bluntly: “The biggest bottleneck in Physical AI is no longer intelligence, it’s experience.”

The choice of locations is strategic. By anchoring its network with two university flagships in Germany, NEURA positions the country as the backbone of Europe’s physical AI infrastructure. Ulrich Rüdiger, rector of RWTH, framed the partnership as “securing Germany’s leading position in the global AI competition.”

This message arrives at a politically charged time. NEURA presents itself as building Physical AI “from Europe,” a sovereignty-driven stance as both the United States and China pour massive resources into humanoid robotics. The company has the financial firepower to back this ambition. Its $1.4 billion Series C funding round, completed in June, attracted investments from NVIDIA, Qualcomm, Amazon, and Bosch.

There are, however, significant caveats. The vision remains largely on paper. Most of the gyms do not yet exist, and the flagship facilities will not open until late this year at the earliest. The model also requires industrial partners to contribute valuable proprietary knowledge, though NEURA insists each partner retains full control over its own data.

The deeper question is more fundamental. Humanoid robots are still largely a demo-stage bet on physical AI, and training reps in a gym are not the same as reliable, consistent work on a production line. But if experience truly is the competitive moat, then the company that builds the gyms may well own the ground on which the future of robotics is learned.