World Labs published a technical post today, July 28, 2026, announcing that robots trained entirely inside a digital simulation ran continuously on real hardware for an hour without human intervention, using its Real-to-Sim-to-Real (R2S2R) framework. This is the first concrete benchmark from the spatial-intelligence company since its $1 billion Series B raised in February 2026 at a $5 billion valuation, bringing total capital raised to $1.23 billion. The paper challenges the robotics field's consensus assumption that simulation can supplement but not replace physical trials. World Labs argues the dominant approach of domain randomization starts in the wrong place, citing a 2026 Annual Reviews survey confirming most simulators "assume perfectly rigid bodies and joints," making deformable object tasks the hardest category for conventional sim-to-real transfer. The R2S2R framework solves the upstream problem first by making the simulation accurate enough. The first phase, Real-to-Sim (R2S), starts with a physical robot task, capturing the robot, sensors, environment, objects, and task demonstrations, then reconstructing all of that as an interactive simulation targeting both visual fidelity and physical fidelity.