Integration of neuroscience and advanced robotics: A paradigm shift for the next generation of intelligent robots

Jinlai Sun1, Rongrong Wang1, Qingsong Wang1, Hongjian Yao1, Qiyue Mu1, Ng Ka Wei2, Hui Yu1, *

1Department of Biomedical Engineering, Medical School of Tianjin University, Tianjin, China;

2DATA4AI Private Limited, Singapore, Singapore

*Correspondence to: Hui Yu, DEng, yuhui@tju.edu.cn.

Intelligent robots are evolving from automated tools that mechanically execute preprogrammed tasks toward a more human-centric paradigm that can adapt Finally, research into to complex, unstructured environments. The core driving force behind this transformation stems from the deep interdisciplinary integration of neuroscience and robotics. This review elaborates on how neuroscience is reshaping the fundamental architecture and developmental trajectory of intelligent robots. We focus on three cutting-edge research areas: brain-computer interface-based human–robot interaction, cognitive state monitoring via neural signal decoding, and neuroscience-inspired perceptual decision-making algorithms coupled with environmental embodied intelligence. In the field of brain–computer interfaces, we analyze how real-time robotic control through brain signals enables more natural and

intuitive human-robot interactions. For cognitive state monitoring, we introduce how neural signal decoding technologies empower robots to perceive users’

psychological states, such as emotion, attention and fatigue, and further adjust their behaviors and feedback accordingly. Finally, research into neuroscience-inspired perceptual decision-making algorithms has enabled robots to perceive the environment and make autonomous decisions in a human-like manner, thereby responding to uncertain, complex, and dynamically changing environments. This work will provide a theoretical framework for the next stage in the evolution of intelligent robots and demonstrate the enormous potential to integrate neuroscience and robotics technologies.

Key Words: brain–computer interface (BCI); cognitive science; embodied intelligence; hybrid electroencephalography; motor imagery (MI); neurorobotics; shared control; spiking neural network (SNN); steady-state visual evoked potential (SSVEP); γ rhythm synchronization