The goal of this project is to develop and validate a radically new model for robotics training — one that breaks from the traditional, linear approach still dominant today. Collaborative robotics training often requires centralized sessions, shared hardware, and instructor-led interaction. Simulators exist, but they’re usually disconnected from real-world setups and offer limited feedback, realism, or transferability. This fragmented pipeline slows learning, increases costs, and excludes many companies—especially SMEs—from building in-house automation skills.We propose a phygital (physical + digital) training system that combines immersive, browser-based simulation with guided physical robot interaction. Learners begin with a high-fidelity digital twin of S3C’s cobotic base cell (CBC), allowing them to safely explore programming, motion planning, and safety-critical tasks—on their own device, without installation or risk. Once prepared, they transition to the physical robot in a tightly coupled session, using the same interface and tasks—minimizing confusion, maximizing confidence.This is not a marginal upgrade; it’s a structural shift: from instructor-led to gamified learning; from disconnected stages to a single, seamless experience. Even when learners physically engage with the robot, they arrive pre-trained and ready, reducing their time on site and increasing safety and efficiency. If successful, this project will lay the groundwork for a distributed, scalable robotics training model—one that supports national upskilling goals, lowers access barriers, and positions Switzerland as a leader in future-ready, modular workforce transformation.