Discover multimodal foundation models that integrate vision, language, and action to enable robots to perceive, understand instructions, and execute physical tasks.