RoboCasa365: A Large-Scale Simulation Framework for Training and Benchmarking Generalist Robots

Researchers have released RoboCasa365, a large-scale simulation benchmark featuring 365 household tasks across 2,500 kitchen environments with over 600 hours of human demonstration data. The platform is designed to train and evaluate generalist robots for everyday tasks, providing insights into factors affecting robot performance and generalization capabilities.

• →RoboCasa365 introduces 365 everyday household tasks across 2,500 diverse kitchen environments for robot training.
• →The platform includes over 600 hours of human demonstration data and 1600+ hours of synthetic data.
• →The benchmark supports multi-task learning, robot foundation model training, and lifelong learning evaluations.
• →Experiments revealed key insights about how task diversity, dataset scale, and environment variation affect robot generalization.
• →This represents one of the most comprehensive resources available for studying generalist robot policies.