Toward Real-World Spatial Intelligence: Integrated Perception and Understanding of Humans, Robots, and Environments

Recent advances in robotics and artificial intelligence are driving a transition from isolated perception and control modules toward integrated spatial intelligence in real-world environments. In healthcare, industrial, and service scenarios, intelligent systems are increasingly expected to perceive and understand humans, robots, objects, and surrounding spaces in a unified manner, while capturing behaviors, intentions, affective states, and contextual dynamics. The rapid development of distributed visual sensing, multimodal learning, and intelligent control has created new opportunities to build human-centered intelligent spaces where perception, interaction, and action are tightly coupled. However, major challenges remain in robust perception under occlusion and noise, multimodal fusion, cross-scene generalization, real-time coordination, and practical deployment in dynamic environments. Addressing these challenges requires new frameworks for integrated perception and understanding across humans, robots, and environments.This Research Topic aims to provide a focused forum for theories, algorithms, and systems that advance integrated perception and understanding of humans, robots, and environments toward real-world spatial intelligence. We seek contributions that move beyond single-task perception and support holistic sensing, reasoning, and interaction in complex physical spaces, particularly through distributed visual sensing, multimodal intelligence, and embodied decision-making.Of particular interest are studies addressing human activity and behavior understanding, object and scene perception, visual control, intention inference, emotion-aware interaction, multimodal agents, embodied intelligence, and context-aware reasoning. We welcome contributions spanning methodological innovation, system integration, and application-oriented validation, provided they offer clear scientific insight and practical value. By bringing together researchers from robotics, computer vision, multimodal AI, and intelligent systems, this Research Topic aims to promote interdisciplinary advances in spatial intelligence for next-generation healthcare, industrial, and service applications.This Research Topic welcomes original research articles, review papers, perspective articles, and methodological studies related to integrated perception and understanding in human-robot-environment systems. Relevant topics include, but are not limited to, human activity recognition, behavior understanding, object detection, scene understanding, distributed visual sensing, multimodal perception and fusion, visual servoing and control, intention understanding, affective computing, multimodal agents, embodied intelligence, human-aware robot interaction, and context-aware intelligence in complex environments. Submissions addressing robustness, interpretability, efficiency, generalization, and real-world deployment are particularly encouraged.Application domains may include smart healthcare, rehabilitation, assistive robotics, intelligent manufacturing, service robotics, and other human-centered intelligent spaces. Manuscripts should demonstrate methodological novelty, technical significance, or convincing evidence of real-world applicability and translational impact. Topic editors: JIAHUI YU, Hongwei Gao, Haibin Cai, Bangli Liu Source: https://www.frontiersin.org/research-topics/80523/toward-real-world-spatial-intelligence-integrated-perception-and-understanding-of-humans-robots-and-environments