World Labs turns one real-world robot task into thousands of simulated variations for training

World Labs has introduced a simulation engine designed to train robotic controllers by generating thousands of virtual variations from a single real-world task. By allowing robots to learn in simulated environments, this technology aims to reduce the data requirements and physical testing time currently needed to deploy autonomous systems in complex settings.