Foundation models have demonstrated remarkable success in advancing machine intelligence, and the research community is now rapidly exploring their potential to bridge cybernetics and intelligence by endowing physical agents such as robots, autonomous vehicles, and smart industrial systems with broad generalization capabilities. This Special Issue focuses on the intersection of massive foundation models and embodied intelligence, examining both the current capabilities and limitations of these models in physical contexts and advancing new architectures, training regimes, and system designs to address these gaps.

The scope includes, but is not limited to:

• Vision-Language-Action (VLA) models and end-to-end embodied planning
• World models and simulation-to-reality transfer
• Efficient deployment on resource-constrained and onboard systems
• Zero-shot and few-shot generalization across diverse embodiments and morphologies
• Multimodal processing (vision, language, tactile sensing, force feedback)
• Data-efficient learning and novel reinforcement or post-training techniques
• Spatial and geometric reasoning in autonomous systems
• Safety, explainability, and trustworthiness for real-world model deployments

All manuscripts submitted to the special issue will be subjected to peer review. Prospective authors should submit an electronic copy of their completed manuscript to https://mc03.manuscriptcentral.com/cai with “Special Issue on Foundation Models for Embodied Intelligence” marked in the cover letter.

Important Date

Manuscript Due: June 15, 2026

Guest Editor

Prof. Yansong Tang, Tsinghua Shenzhen International Graduate School, China

E-mail: tang.yansong@sz.tsinghua.edu.cn