The EU Built an XR Training Platform. Robots May Never Look the Same.
The EU Built an XR Training Platform. Robots May Never Look the Same.
On June 9, in Oldenburg, Germany, the MASTER consortium will spend a day showing companies what an open, ROS-native (compatible with the Robot Operating System) XR training platform looks like when 78 organizations have already tried to build on it. That question — whether the demand signal is real enough to outlast the funding — is the whole story.
MASTER is a €8.34 million EU Horizon Europe project that launched in January 2023 and runs through December 2026. The consortium — seven partners across three countries, anchored by the Laboratory for Manufacturing Systems and Automation at the University of Patras — has spent three years building an open extended reality platform for industrial robotics training. The platform, called MASTER Open XR, lets factory workers practice robot programming and safety procedures in a virtual environment before touching real equipment.
When Panagiotis Karagiannis, the project coordinator, talks about why factories still struggle to train workers on robots, he keeps returning to fear. Not the philosophical kind — the visceral kind. "Robots make quick movements and sudden sounds that can be intimidating," he says. His solution wasn't a better manual or a longer orientation. It was a digital environment where trainees could fail without consequence.
That's not a novel story. VR training in manufacturing has been in pilot programs since the early 2010s. What's interesting is what MASTER built underneath it.
The platform runs on VIROO, an enterprise XR environment developed by Spanish VR firm Virtualware. MASTER added three features designed specifically for factory training: code-free robot programming, gaze-based interaction, and what the project calls safe robotic environments — virtual zones equipped with sensors that trigger a speed reduction or protective stop when a human enters the space. The platform is compatible with OpenXR and ROS, the open-source Robot Operating System that serves as the common language for industrial robot communication.
That ROS compatibility is the part worth dwelling on.
ROS is not a product. It's an ecosystem — a set of open-source tools and libraries that robot manufacturers including ABB, KUKA, Fanuc, and Universal Robots have all built around. If a training platform speaks ROS natively, a worker trained on MASTER isn't just learning generic robotics concepts. They're learning the specific vocabulary and workflows they'll encounter on actual factory floors using equipment from major vendors. The training credential has crossover value that a proprietary simulation environment doesn't.
"The platform can really showcase to industry that through collaboration, we can create educational scenarios that deliver good results and provide a supportive environment for training operators and engineers," Karagiannis said in an interview with EE Times.
MASTER tested this theory by opening its platform to third-party developers through two Open Calls. The first, launched in March 2024, received 64 applications and funded 17 projects. The second, launched in April 2025, received 78 applications — a 22 percent increase — and funded 24. Forty-one external projects now sit on top of the MASTER platform, building add-ons that range from voice-prompted spawning tools for factory floor objects to a photo-realistic digital twin of a cargo ship for maritime safety training. Nokia Bell Labs contributed to the add-on developer roster.
The project has produced more than 30 assets — 3D models, safety zones, user interfaces, a virtual assistant. The work is documented in a peer-reviewed paper presented at ACM's VRST 2024 conference, which describes the platform's modules for creating safe workspaces, intuitive robot programming, and eye-tracking interaction.
The final showcase is June 9 in Oldenburg, Germany.
Here's what we don't know: what happens to all of this after December, when the EU funding cycle ends and the consortium disperses. Virtualware has a commercial product in VIROO, and the MASTER platform is built on top of it. The Open Calls funded prototype work — proof of concept, not necessarily production-grade deployment. Whether any industrial manufacturer has signed a commercial licensing agreement with Virtualware, or whether any of the 41 funded projects have been deployed in active factories, is not public information.
The ROS compatibility framing could age well or badly depending on the answer. If VIROO's ROS integration is robust — a real communication layer between the training environment and actual robot controllers — then MASTER built something close to a standard. If it's surface-level, a checkbox that sounds good on a grant application, then the talent pipeline argument collapses. You'd have a training simulator that teaches habits that don't transfer to the shop floor.
EU Horizon projects have a spotty commercial track record. The dissemination budgets that fund press releases and conference papers don't always translate into durable products. MASTER is a research consortium, not a company, and the gap between "developed more than 30 assets" and "operational deployment at scale" is one that many EU projects never close.
What's real is the demand signal. Seventy-eight organizations applied for the second Open Call. That number doesn't lie about the appetite, even if the supply side hasn't caught up yet. The MASTER consortium built the skeleton of something — an open, ROS-native XR training infrastructure that external developers are apparently willing to build on. Whether Virtualware can turn that skeleton into a business, and whether the training outcomes actually translate to factory floor performance, are the questions that will determine whether this was eight million euros well spent or another EU research project that produced good papers and little else.
The final event in Oldenburg should clarify which answer is closer to true.