Topic: Embodied Intelligence & Multimodal Learning for Robotics

Robotics is rapidly advancing toward autonomous operation in unstructured, dynamic environments, driven by two interconnected pillars: embodied intelligence and learning-based methods. Embodied intelligence, which integrates physical and cognitive capabilities, plays a crucial role in achieving robust, real-time decision making and autonomous manipulation. Learning-based methods, incorporating advances in machine learning such as Large Language Models (LLMs), Visual Language Models (VLMs), Multimodal Learning, Reinforcement Learning, and Imitation Learning, significantly enhance the practical abilities of on-site robots.

While embodied intelligence and learning-based methods have each propelled robotics forward, significant gaps persist in their integration, particularly for learning-based manipulation tasks and real-world deployment. This Special Issue seeks to address these gaps by promoting research at the intersection of embodied intelligence and learning-based methods, while exploring novel methodologies, innovative systems, and real-world implementations. We warmly invite authors to submit original research articles and reviews that advance the state of the art in this rapidly evolving field.

Topics of interest include, but are not limited to:

● Robotic learningvia imitation learning, reinforcement learning, and transfer learning;

● Applications of LLMs and VLMs for multimodal sensory fusion (e.g., vision, tactile sensing, and proprioception) in complex manipulation tasks;

● Multimodal perception fusion and its integration intorobotic control systems;

● Human-robot interaction and learning from human feedback;

● Simulation-to-real transfer for multimodal embodied learningin robotics;

● Verifiable and safe embodied intelligence.

Embodied intelligence, multimodal learning, robot learning, human-robot interaction, imitation learning, reinforcement learning