The field of embodied intelligence is moving towards the development of more generalizable and scalable models that can seamlessly integrate perception and action across diverse platforms. Researchers are focusing on enhancing the spatial awareness and reasoning capabilities of Vision-Language Models (VLMs) to enable more human-like intelligence. The introduction of new frameworks and datasets is facilitating the progression from specialized task-oriented systems to more generalist and cognitively capable agents. Noteworthy papers in this regard include: iFlyBot-VLM Technical Report, which introduces a general-purpose Vision-Language Model that bridges the cross-modal semantic gap between environmental perception and robotic motion control. Visual Spatial Tuning, which presents a comprehensive framework to cultivate VLMs with human-like visuospatial abilities. Data Assessment for Embodied Intelligence, which introduces principled data-driven tools to evaluate dataset diversity and learnability. Fundamentals of Physical AI, which elaborates the fundamental principles of physical artificial intelligence from a scientific and systemic perspective.